Geometric appearance defect detection method for solar cell
阅读说明：本技术 一种太阳能电池片几何外观缺陷检测方法 (Geometric appearance defect detection method for solar cell ) 是由 沈满德 邵坤 张法全 周利兵 闵玉瑾 于 2019-10-21 设计创作，主要内容包括：本发明公开了一种太阳能电池片几何外观缺陷检测方法,包含如下步骤：S1、取标准太阳能电池片,根据该标准太阳能电池片规格尺寸和用户要求,创建个性化检测标准；S2、取待检测太阳能电池片,根据待检测太阳能电池片规格尺寸和用户要求,从步骤S1所创建的个性化检测标准中选择与待检测太阳能电池片的型号规格和用户检测要求相一致的个性化检测标准,进行几何外观缺陷检测。本发明的优点有：可实现多种几何外观缺陷的同时检测；即可用于单块太阳能电池片检测,也可用于多块太阳能电池片的同时检测；能有效区分倒角与缺角；根据太阳能电池片规格尺寸和用户要求,可创建个性化检测标准。(The invention discloses a method for detecting geometric appearance defects of a solar cell, which comprises the following steps: s1, taking a standard solar cell, and creating a personalized detection standard according to the specification and the size of the standard solar cell and the user requirement; and S2, taking the solar cell to be detected, selecting an individualized detection standard which is consistent with the model specification and the user detection requirement of the solar cell to be detected from the individualized detection standards created in the step S1 according to the specification and the size of the solar cell to be detected and the user detection requirement, and detecting the geometrical appearance defects. The invention has the advantages that: the simultaneous detection of various geometric appearance defects can be realized; the method can be used for detecting a single solar cell and can also be used for simultaneously detecting a plurality of solar cells; the chamfer and the unfilled corner can be effectively distinguished; according to the specification and the size of the solar cell and the requirements of a user, an individualized detection standard can be created.)
1. A method for detecting geometric appearance defects of a solar cell is characterized by comprising the following steps: comprises the following steps:
s1, taking a standard solar cell, and creating a personalized detection standard according to the specification and the size of the standard solar cell and the user requirement;
the specification size of the solar cell slice comprises the length and width size and the chamfer size of the solar cell slice; the user requirement refers to a control standard when the user detects various defects; the standard solar cell is a solar cell piece with the size meeting the requirement and without unfilled corners, broken edges, holes and crack defects; the individualized detection standard is set according to the specification and the size of a standard solar cell and specific requirements of a user, and comprises an allowable range of the length and the width of the solar cell, an allowable range of the size of a chamfer, and judgment values of various defects such as unfilled corners, broken edges, holes, cracks and the like;
and S2, taking the solar cell to be detected, selecting an individualized detection standard which is consistent with the model specification and the user detection requirement of the solar cell to be detected from the individualized detection standards created in the step S1 according to the specification and the size of the solar cell to be detected and the user detection requirement, and detecting the geometrical appearance defects.
2. The method for detecting the geometric appearance defects of the solar cell according to claim 1, wherein the method comprises the following steps: the step S1 specifically includes the following steps:
s1-1, naming the personalized detection standard;
the naming personalized detection standard specifically refers to that different detection standards are adopted according to different specifications of solar cells and different user requirements, and the detection standards are distinguished through different naming;
s1-2, collecting images of the standard solar cell;
the standard solar cell is a solar cell with the size meeting the requirement and without defects of unfilled corners, broken edges, holes and cracks;
the image is generated by a standard solar cell under the condition of a backlight source, the area of the standard solar cell in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s1-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the standard solar cell image acquired in the step S1-2 by using algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to the step of performing binarization processing on a standard solar cell image subjected to image preprocessing by using a segmentation threshold, wherein the segmentation threshold is 128;
s1-4, selecting all standard solar cell areas and sorting according to positions;
the standard solar cell region is obtained after the image segmentation in the step S1-3, the region obtained after the image segmentation in the step S1-3 contains the standard solar cell region or contains the standard solar cell region and an interference region, and when the region obtained after the image segmentation in the step S1-3 contains the standard solar cell region and the interference region, the standard solar cell region needs to be extracted according to the characteristics of the length-width ratio, the rectangular degree, the coordinate range and the like of the solar cell, and the interference region is removed;
the sorting according to positions specifically means that when the region obtained after the image segmentation in the step S1-3 includes a plurality of standard solar cell regions, sorting is required according to the respective positions, where the positions refer to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each standard solar cell region;
s1-5, selecting standard solar cell areas one by one;
specifically, the step of selecting the standard solar cell areas one by one includes the steps of sorting the standard solar cell areas one by one according to positions in the step S1-4, and selecting one standard solar cell area one by one according to the order;
s1-6, inputting the length and width size and the chamfer size of the selected standard solar cell, and setting control parameters such as a defect area threshold, a defect length threshold, a cell length and width size allowable range, a defect number allowable value and the like;
the length and width of the solar cell is specifically the length and width of a rectangle externally connected with a standard solar cell, and the chamfer size is specifically two side lengths of a straight chamfer of the standard solar cell;
the defect area threshold and the defect length threshold are set according to specific requirements of a user and are used for defect filtering;
when the defect area of the selected standard solar cell to be detected is larger than the set defect area threshold value, determining that the defect area is a defect; otherwise, it is not a defect;
when the maximum diameter of the selected standard solar cell to be detected is larger than a set defect length threshold value, determining that the defect is a defect; otherwise, it is not a defect;
the length and width size allowable range of the cell slice refers to the length and width size allowable range of a circumscribed rectangle of a standard solar cell;
when the length and width of the selected standard solar cell slice circumscribed rectangle to be detected exceeds the set range, the standard solar cell slice is an unqualified product;
the defect number allowable value is the maximum value of the sum of the four types of defects of unfilled corners, broken edges, holes and cracks, and the standard solar cell to be detected is qualified when the sum of the four types of defects of the selected standard solar cell to be detected is less than the set value; otherwise, the product is unqualified;
s1-7, extracting four sides of the selected standard solar cell area, calculating the intersection points of the four sides, and obtaining the external quadrangle of the standard solar cell area according to the four intersection points;
the four sides of the standard solar cell area specifically refer to the edge contour of the standard solar cell area extracted by using an edge detection algorithm, and after smooth fitting, the longest line segments on the four sides are respectively selected;
the intersection point of the four sides specifically refers to an intersection point between straight lines where four longest line segments are located in the edge outline of the extracted standard solar cell area, and a quadrangle determined by the four intersection points is an external quadrangle of the selected standard solar cell area to be obtained;
s1-8, making a difference between the circumscribed quadrangle of the selected standard solar cell area and the selected standard solar cell area, and obtaining four chamfers of the selected standard solar cell;
the four chamfers are obtained by subtracting the selected standard solar cell area from the circumscribed quadrangle of the selected standard solar cell area obtained in the step S1-7, and the chamfer size and the chamfer area of each chamfer are used for judging the corner defect;
s1-9, calculating the length and width scale factor of the circumscribed quadrangle;
the length-width scaling factor is obtained by dividing the length-width dimension value of the standard solar cell slice input in the step S1-6 by the long axis and short axis pixel values of the circumscribed quadrangle of the selected standard solar cell slice region obtained in the step S1-7;
s1-10, calculating the geometric parameters of the selected standard solar cell area by using the length-width scale factor, and storing the geometric parameters into the created personalized detection standard;
the geometric parameters comprise the area, the squareness, the four chamfer sizes and the like, and are specifically the geometric parameters comprising the area, the squareness, the four chamfer sizes and the like obtained after converting pixel units into physical size units through length-width scale factors;
and S1-11, repeating the steps S1-5 to S1-10 until all the standard solar cell areas are processed.
3. The method for detecting the geometric appearance defects of the solar cell according to claim 2, wherein the method comprises the following steps: the step S2 specifically includes the following steps:
s2-1, selecting a personalized detection standard;
the individualized detection standard is specifically the individualized detection standard created in step S1, and the selected individualized detection standard is consistent with the model specification of the solar cell to be detected and the user detection requirement;
the personalized detection standard comprises control parameters such as cell length and width scale factors, defect area thresholds, defect length thresholds, cell areas, cell chamfer sizes, cell length and width size allowable ranges and defect number allowable values, and whether the solar cell to be detected is qualified is judged according to the control parameters;
the length-width scaling factor refers to a conversion relation between a pixel unit and a physical size unit, and is used for converting control parameters such as a defect area, a defect length, a cell area, a cell chamfer size, a cell length-width size and the like from the pixel unit to the physical size unit so as to realize defect judgment;
s2-2, collecting an image of the solar cell to be detected;
the solar cell to be detected is a solar cell which may have geometric appearance defects such as size deviation, unfilled corners, edge breakage, holes or cracks;
the image is generated by the solar cell to be detected under the backlight source condition, the area of the solar cell to be detected in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s2-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the solar cell image to be detected acquired in the step S2-2 by adopting algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to the step of performing binarization processing on a solar cell image to be detected after image preprocessing by using a segmentation threshold, wherein the segmentation threshold is 128;
s2-4, selecting all solar cell areas to be detected, and sorting according to positions;
the solar cell region to be detected specifically refers to a region obtained after image segmentation in the step S2-3, the region obtained after image segmentation in the step S2-3 includes the solar cell region to be detected or includes the solar cell region to be detected and an interference region, and when the region obtained after image segmentation in the step S2-3 includes the solar cell region to be detected and the interference region, the solar cell region to be detected needs to be extracted according to the characteristics of the length-width ratio, the rectangular degree, the coordinate range and the like of the solar cell, and the interference region is removed;
the sorting according to the position refers to sorting according to the respective position when the region obtained after the image segmentation in the step S2-3 contains a plurality of solar cell regions to be detected, wherein the position refers to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each solar cell region to be detected;
s2-5, selecting the solar cell areas to be detected one by one;
specifically, the step of selecting the solar cell areas to be detected one by one is to select a certain solar cell area to be detected one by one in sequence after sorting according to positions in the step S2-4;
s2-6, calculating the area, the rectangular degree and the length and the width of the minimum circumscribed rectangle of the selected solar cell area to be detected, and converting the area, the rectangular degree and the length and the width into geometric parameters;
the area of the solar cell area to be detected refers to the number of pixels of the area obtained after the binarization processing of the solar cell to be detected;
the rectangularity refers to the similarity between a solar cell area to be detected and a rectangle;
the length and width of the minimum circumscribed rectangle refers to the length and width of the minimum circumscribed rectangle obtained by fitting the solar cell to be detected at any inclination angle;
the geometric parameters are length and width scale factors led in by using the personalized detection standards, and the area, the rectangular degree and the length and width of the minimum external rectangle of the solar cell to be detected are converted into physical size units from pixel units so as to be compared with control parameters in the selected personalized detection standards;
s2-7, performing primary judgment to judge whether the geometric parameters meet the requirements;
if not, determining that the detection is not qualified, and ending the detection;
if so, performing subsequent detection;
the primary judgment specifically refers to comparing the geometric parameters obtained in the step S2-6 with the control parameters in the personalized detection standard selected in the step S2-1 to judge whether the geometric parameters are within an allowable range;
s2-8, extracting four sides of the selected solar cell area to be detected, calculating intersection points of the four sides, and obtaining an external quadrangle of the solar cell area to be detected according to the four intersection points;
the four edges of the selected solar cell area to be detected are extracted and are subjected to subsequent detection after the first detection in the step S2-7 is qualified, specifically, the edge contour of the selected solar cell area to be detected is extracted by using an edge detection algorithm, and after smooth fitting, the longest line segments on the four edges are respectively selected;
the intersection points of the four sides are intersection points between straight lines where the four longest line segments are located in the edge outline of the selected solar cell area to be detected, and a quadrangle determined by the four intersection points is an external quadrangle of the selected solar cell area to be detected to be obtained;
s2-9, cutting the circumscribed quadrangle of the selected solar cell area to be detected by utilizing the four chamfer sizes and the length and width scale factors in the selected personalized detection standard;
the cutting of the external quadrangle of the solar cell area to be detected specifically means that four chamfers are subtracted from the external quadrangle of the solar cell area to be detected to obtain the solar cell area to be detected without the chamfers, so that the chamfers are prevented from being misjudged as unfilled corners;
s2-10, filling and closing the cut area by using morphology;
the cut region refers to a solar cell region to be detected which is processed in the step S2-9 and does not contain four chamfers;
filling holes and cracks in the region through filling operation, selecting a structural factor consistent with the direction of an external quadrangle, closing a broken edge gap on the edge of the region through closing operation, wherein the external quadrangle after filling and closing operation is a completely convex polygon without four chamfers and any hole cracks and broken edge gaps;
s2-11, making a difference between the morphologically processed region and the selected solar cell region to be detected to obtain a defect candidate region;
the morphologically processed region is specifically an outward convex polygonal region obtained after the processing in the step S2-10, and the selected to-be-detected cell region is specifically a certain to-be-detected solar cell region selected in the step S2-5; the difference is specifically obtained by subtracting the two areas; the defect candidate area is obtained by subtracting the two areas to obtain all defect candidate areas of unfilled corners, broken edges, holes and cracks;
s2-12, judging the alternative areas by using the control parameters in the personalized detection standard, and screening out real defects;
the control parameters specifically refer to parameters such as a defect area threshold value, a defect length threshold value and the like which are imported when the personalized detection standard is selected in the step S2-1;
the candidate region is the defect candidate region obtained in the step S2-11, and includes all unfilled corners, broken edges, holes and crack defects;
s2-13, classifying the defects according to the positions and the shape characteristic values of the defects;
the defect classification means that the defects are classified into one of unfilled corners, broken edges, holes and cracks respectively according to the positions and the shape characteristic values of the defects;
s2-14, judging whether the defect number exceeds the limit;
judging whether the defect number exceeds the limit specifically by comparing the defect number of the solar cell to be detected with a defect number allowable value in the personalized detection standard selected in the step S2-1, judging whether the defect number exceeds the limit, and if the defect number exceeds the limit, judging that the defect number is unqualified; if not, judging the product is qualified;
the defect number refers to the sum of four types of defects, namely corner defect, edge breakage, hole and crack;
s2-15, repeating the steps S2-5 to S2-14 until all the solar cell areas to be detected are processed;
s2-16, storing and outputting the detection result;
the step of storing the detection result specifically refers to storing the detection result, including information such as an actual measurement value, an allowable value range, a defect number, unqualified reasons and the like, into a database file;
the output specifically refers to sending the detection result to a relevant receiving end in a Modbus communication protocol or other modes according to an agreed protocol.
Technical Field
The invention relates to the technical field of solar cells, in particular to a method for detecting geometric appearance defects of a solar cell.
Background
The conventional fossil energy cannot be regenerated, and the requirement of human on the energy for continuous development cannot be met; in addition, such energy consumption causes the environmental problems to be worsened continuously, and the treatment cost is higher and higher. Therefore, the popularization of clean energy is increasingly urgent, and solar energy is one of important clean energy and is highly valued by countries in the world. In recent years, the proportion of solar power generation to energy is steadily increased.
The photovoltaic module is an important device for converting solar power generation light energy into electric energy, once the photovoltaic module is installed, the normal service life of the photovoltaic module can generally reach 30 years, so that the quality of the photovoltaic module is ensured, and the photovoltaic module has important significance for improving the solar power generation efficiency and realizing the flat price internet surfing.
The photovoltaic module is formed by series welding and packaging a plurality of battery pieces, so that the quality of the battery pieces plays a decisive role in the quality of the photovoltaic module. The battery piece is mainly made of silicon materials, has fragility, and is easy to have geometrical appearance defects in the processing and series welding process, thereby influencing the quality of components.
The geometric appearance defects of the battery piece mainly comprise: size deviation, unfilled corners, edge breakage, holes and cracks. Therefore, before series welding of the battery piece, geometric appearance defect detection is carried out, unqualified products are removed, the fact that the battery piece entering the series welding has no geometric appearance defect is guaranteed, and the method has important significance for improving the quality of the assembly.
Through searching domestic patents, the patents related to the detection of geometric appearance defects of the battery piece mainly include:
1. chinese patent application No. 201210175496.7, publication No. CN103454280A, and lielongqiang et al propose "a method for determining damage of solar cell". The method adopts the steps that after the battery piece image is rotated by 180 degrees, the battery piece image is compared with an original image to obtain the damage condition. The method has the following defects: the cell is required to have symmetry; the method can not be used for detecting holes and cracks; chamfering of the battery piece is not considered during corner defect detection; the method can not be used for simultaneously detecting a plurality of battery slices.
2. Chinese patent application No. 201810424920.4, publication No. CN108596903A, ear wearing, etc. provides a method for detecting defects of black edges and fragments of solar cells. The method is suitable for EL image detection of the assembly, and cannot be used for visible light image breakage detection of the battery piece.
In summary, the prior art mainly has the following disadvantages:
1. the defect detection type is single, and the comprehensive detection of all defects of the geometric appearance defects cannot be realized.
2. Only a single cell can be detected, and the image cannot contain a plurality of cells.
3. The detection precision is low, and defects smaller than small can not be detected.
Disclosure of Invention
Aiming at the steps of the background technology, the invention aims to provide a solar cell geometric appearance defect detection method which can be used for detecting a single cell and a plurality of cells and can realize comprehensive detection of various geometric appearance defects.
In order to achieve the purpose, the invention adopts the technical scheme that: a method for detecting geometric appearance defects of a solar cell comprises the following steps:
s1, taking a standard solar cell, and creating a personalized detection standard according to the specification and the size of the standard solar cell and the user requirement;
the specification size of the solar cell slice comprises the length and width size and the chamfer size of the solar cell slice; the user requirement refers to a control standard when the user detects various defects; the standard solar cell is a solar cell with the size meeting the requirement and without unfilled corners, broken edges, holes and crack defects; the individualized detection standard is set according to the specification and the size of the solar cell and the specific requirements of a user, and comprises an allowable range of the length and the width of the solar cell, an allowable range of the size of a chamfer and judgment values of various defects such as unfilled corners, broken edges, holes, cracks and the like;
and S2, taking the solar cell to be detected, selecting an individualized detection standard which is consistent with the model specification and the user detection requirement of the solar cell to be detected from the individualized detection standards created in the step S1 according to the specification and the size of the solar cell to be detected and the user detection requirement, and detecting the geometrical appearance defects.
In the above detection method, the step S1 specifically includes the following steps:
s1-1, naming the personalized detection standard;
the naming personalized detection standard specifically refers to that different detection standards are adopted according to different specifications of solar cells and different user requirements, and the detection standards are distinguished through different naming;
s1-2, collecting images of the standard solar cell;
the standard solar cell is a solar cell with the size meeting the requirement and without defects of unfilled corners, broken edges, holes and cracks;
the image is generated by a standard solar cell under the condition of a backlight source, the area of the standard solar cell in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s1-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the standard solar cell image acquired in the step S1-2 by using algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to the step of performing binarization processing on a standard solar cell image subjected to image preprocessing by using a segmentation threshold, wherein the segmentation threshold is 128;
s1-4, selecting all standard solar cell areas and sorting according to positions;
the standard solar cell region is obtained after the image segmentation in the step S1-3, the region obtained after the image segmentation in the step S1-3 contains the standard solar cell region or contains the standard solar cell region and an interference region, and when the region obtained after the image segmentation in the step S1-3 contains the standard solar cell region and the interference region, the standard solar cell region needs to be extracted according to the characteristics of the length-width ratio, the rectangular degree, the coordinate range and the like of the solar cell, and the interference region is removed;
the sorting according to positions specifically means that when the region obtained after the image segmentation in the step S1-3 includes a plurality of standard solar cell regions, sorting is required according to the respective positions, where the positions refer to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each standard solar cell region;
s1-5, selecting standard solar cell areas one by one;
specifically, the step of selecting the standard solar cell areas one by one includes the steps of sorting the standard solar cell areas one by one according to positions in the step S1-4, and selecting one standard solar cell area one by one according to the order;
s1-6, inputting the length and width size and the chamfer size of the selected standard solar cell, and setting control parameters such as a defect area threshold, a defect length threshold, a cell length and width size allowable range, a defect number allowable value and the like;
the length and width of the solar cell is specifically the length and width of a rectangle externally connected with a standard solar cell, and the chamfer size is specifically two side lengths of a straight chamfer of the standard solar cell;
the defect area threshold and the defect length threshold are set according to specific requirements of a user and are used for defect filtering;
when the defect area of the selected standard solar cell to be detected is larger than the set defect area threshold value, determining that the defect area is a defect; otherwise, it is not a defect;
when the maximum diameter of the selected standard solar cell to be detected is larger than a set defect length threshold value, determining that the defect is a defect; otherwise, it is not a defect;
the length and width size allowable range of the cell slice refers to the length and width size allowable range of a circumscribed rectangle of a standard solar cell;
when the length and width of the selected standard solar cell slice circumscribed rectangle to be detected exceeds the set range, the standard solar cell slice is an unqualified product;
the defect number allowable value is the maximum value of the sum of the four types of defects of unfilled corners, broken edges, holes and cracks, and the standard solar cell to be detected is qualified when the sum of the four types of defects of the selected standard solar cell to be detected is less than the set value; otherwise, the product is unqualified;
s1-7, extracting four sides of the selected standard solar cell area, calculating the intersection points of the four sides, and obtaining the external quadrangle of the standard solar cell area according to the four intersection points;
the four sides of the standard solar cell area specifically refer to the edge contour of the standard solar cell area extracted by using an edge detection algorithm, and after smooth fitting, the longest line segments on the four sides are respectively selected;
the intersection point of the four sides specifically refers to an intersection point between straight lines where four longest line segments are located in the edge outline of the extracted standard solar cell area, and a quadrangle determined by the four intersection points is an external quadrangle of the selected standard solar cell area to be obtained;
s1-8, making a difference between the circumscribed quadrangle of the selected standard solar cell area and the selected standard solar cell area, and obtaining four chamfers of the selected standard solar cell;
the four chamfers are obtained by subtracting the selected standard solar cell area from the circumscribed quadrangle of the selected standard solar cell area obtained in the step S1-7, and the chamfer size and the chamfer area of each chamfer are used for judging the corner defect;
s1-9, calculating the length and width scale factor of the circumscribed quadrangle;
the length-width scaling factor is obtained by dividing the length-width dimension value of the standard solar cell slice input in the step S1-6 by the long axis and short axis pixel values of the circumscribed quadrangle of the selected standard solar cell slice region obtained in the step S1-7;
s1-10, calculating the geometric parameters of the selected standard solar cell area by using the length-width scale factor, and storing the geometric parameters into the created personalized detection standard;
the geometric parameters comprise the area, the squareness, the four chamfer sizes and the like, and are specifically the geometric parameters comprising the area, the squareness, the four chamfer sizes and the like obtained after converting pixel units into physical size units through length-width scale factors;
and S1-11, repeating the steps S1-5 to S1-10 until all the standard solar cell areas are processed.
In the above detection method, the step S2 specifically includes the following steps:
s2-1, selecting a personalized detection standard;
the individualized detection standard is specifically the individualized detection standard created in step S1, and the selected individualized detection standard is consistent with the model specification of the solar cell to be detected and the user detection requirement;
the personalized detection standard comprises control parameters such as cell length and width scale factors, defect area thresholds, defect length thresholds, cell areas, cell chamfer sizes, cell length and width size allowable ranges and defect number allowable values, and whether the solar cell to be detected is qualified is judged according to the control parameters;
the length-width scaling factor refers to a conversion relation between a pixel unit and a physical size unit, and is used for converting control parameters such as a defect area, a defect length, a cell area, a cell chamfer size, a cell length-width size and the like from the pixel unit to the physical size unit so as to realize defect judgment;
s2-2, collecting an image of the solar cell to be detected;
the solar cell to be detected is a solar cell which may have geometric appearance defects such as size deviation, unfilled corners, edge breakage, holes or cracks;
the image is generated by the solar cell to be detected under the backlight source condition, the area of the solar cell to be detected in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s2-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the solar cell image to be detected acquired in the step S2-2 by adopting algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to the step of performing binarization processing on a solar cell image to be detected after image preprocessing by using a segmentation threshold, wherein the segmentation threshold is 128;
s2-4, selecting all solar cell areas to be detected, and sorting according to positions;
the solar cell region to be detected specifically refers to a region obtained after image segmentation in the step S2-3, the region obtained after image segmentation in the step S2-3 includes the solar cell region to be detected or includes the solar cell region to be detected and an interference region, and when the region obtained after image segmentation in the step S2-3 includes the solar cell region to be detected and the interference region, the solar cell region to be detected needs to be extracted according to the characteristics of the length-width ratio, the rectangular degree, the coordinate range and the like of the solar cell, and the interference region is removed;
the sorting according to the position refers to sorting according to the respective position when the region obtained after the image segmentation in the step S2-3 contains a plurality of solar cell regions to be detected, wherein the position refers to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each solar cell region to be detected;
s2-5, selecting the solar cell areas to be detected one by one;
specifically, the step of selecting the solar cell areas to be detected one by one is to select a certain solar cell area to be detected one by one in sequence after sorting according to positions in the step S2-4;
s2-6, calculating the area, the rectangular degree and the length and the width of the minimum circumscribed rectangle of the selected solar cell area to be detected, and converting the area, the rectangular degree and the length and the width into geometric parameters;
the area of the solar cell area to be detected refers to the number of pixels of the area obtained after the binarization processing of the solar cell to be detected;
the rectangularity refers to the similarity between a solar cell area to be detected and a rectangle;
the length and width of the minimum circumscribed rectangle refers to the length and width of the minimum circumscribed rectangle obtained by fitting the solar cell to be detected at any inclination angle;
the geometric parameters are length and width scale factors led in by using the personalized detection standards, and the area, the rectangular degree and the length and width of the minimum external rectangle of the solar cell to be detected are converted into physical size units from pixel units so as to be compared with control parameters in the selected personalized detection standards;
s2-7, performing primary judgment to judge whether the geometric parameters meet the requirements;
if not, determining that the detection is not qualified, and ending the detection;
if so, performing subsequent detection;
the primary judgment specifically refers to comparing the geometric parameters obtained in the step S2-6 with the control parameters in the selected personalized detection standard to judge whether the geometric parameters are within an allowable range;
s2-8, extracting four sides of the selected solar cell area to be detected, calculating intersection points of the four sides, and obtaining an external quadrangle of the solar cell area to be detected according to the four intersection points;
the four edges of the selected solar cell area to be detected are extracted and are subjected to subsequent detection after the first detection in the step S2-7 is qualified, specifically, the edge contour of the selected solar cell area to be detected is extracted by using an edge detection algorithm, and after smooth fitting, the longest line segments on the four edges are respectively selected;
the intersection points of the four sides are intersection points between straight lines where the four longest line segments are located in the edge outline of the selected solar cell area to be detected, and a quadrangle determined by the four intersection points is an external quadrangle of the selected solar cell area to be detected to be obtained;
s2-9, cutting the circumscribed quadrangle of the selected solar cell area to be detected by utilizing the four chamfer sizes and the length and width scale factors in the selected personalized detection standard;
the cutting of the external quadrangle of the solar cell area to be detected specifically means that four chamfers are subtracted from the external quadrangle of the solar cell area to be detected to obtain the solar cell area to be detected without the chamfers, so that the chamfers are prevented from being misjudged as unfilled corners;
s2-10, filling and closing the cut area by using morphology;
the cut region refers to a solar cell region to be detected which is processed in the step S2-9 and does not contain four chamfers;
filling holes and cracks in the region through filling operation, selecting a structural factor consistent with the direction of an external quadrangle, closing a broken edge gap on the edge of the region through closing operation, wherein the external quadrangle after filling and closing operation is a completely convex polygon without four chamfers and any hole cracks and broken edge gaps;
s2-11, making a difference between the morphologically processed region and the selected solar cell region to be detected to obtain a defect candidate region;
the morphologically processed region is specifically an outward convex polygonal region obtained after the processing in the step S2-10, and the selected to-be-detected cell region is specifically a certain to-be-detected solar cell region selected in the step S2-5; the difference is specifically obtained by subtracting the two areas; the defect candidate area is obtained by subtracting the two areas to obtain all defect candidate areas of unfilled corners, broken edges, holes and cracks;
s2-12, judging the alternative areas by using the control parameters in the personalized detection standard, and screening out real defects;
the control parameters specifically refer to parameters such as a defect area threshold value, a defect length threshold value and the like which are imported when the personalized detection standard is selected in the step S2-1;
the candidate region is the defect candidate region obtained in the step S2-11, and includes all unfilled corners, broken edges, holes and crack defects;
s2-13, classifying the defects according to the positions and the shape characteristic values of the defects;
the defect classification means that the defects are classified into one of unfilled corners, broken edges, holes and cracks respectively according to the positions and the shape characteristic values of the defects;
s2-14, judging whether the defect number exceeds the limit;
judging whether the defect number exceeds the limit specifically by comparing the defect number of the solar cell to be detected with a defect number allowable value in the personalized detection standard selected in the step S2-1, judging whether the defect number exceeds the limit, and if the defect number exceeds the limit, judging that the defect number is unqualified; if not, judging the product is qualified;
the defect number refers to the sum of four types of defects, namely corner defect, edge breakage, hole and crack;
s2-15, repeating the steps S2-5 to S2-14 until all the solar cell areas to be detected are processed;
s2-16, storing and outputting the detection result;
the step of storing the detection result specifically refers to storing the detection result, including information such as an actual measurement value, an allowable value range, a defect number, unqualified reasons and the like, into a database file;
the output specifically refers to sending the detection result to the receiving end in a Modbus communication protocol or other modes according to an agreed protocol.
Compared with the prior art, the invention has the advantages that:
1. the method can realize the simultaneous detection of various geometric appearance defects, and the defect types comprise size deviation, unfilled corners, edge breakage, holes and cracks.
2. The method can be used for detecting a single solar cell and can also be used for simultaneously detecting a plurality of solar cells.
3. The detection precision is high, and the minimum size of the defect can reach 0.1 mm.
4. The device is suitable for detecting various solar cell pieces, including single crystal cell pieces and polycrystalline cell pieces, and can effectively distinguish chamfers from unfilled corners.
5. According to the specification and the size of the solar cell and the requirements of a user, an individualized detection standard can be created.
The innovation points of the invention are as follows: compared with the existing solar cell appearance detection method, the invention creatively provides the advantages that the simultaneous detection of various appearance defects and the simultaneous detection of a plurality of solar cells are realized by establishing the personalized detection standard, and the solar cell appearance detection method is compatible with the solar cells of various specifications, has higher detection precision and can effectively improve the quality of the solar cells for series welding.
Drawings
FIG. 1 is a process flow diagram of a method for detecting geometric appearance defects of a solar cell according to the present invention;
FIG. 2 is a flowchart illustrating a first stage of the method for detecting geometric defects of a solar cell according to the present invention;
FIG. 3 is a flowchart illustrating a second stage of the method for detecting geometric defects of solar cells according to the present invention;
FIG. 4 is an image extracted when the method for detecting geometric appearance defects of a solar cell according to the present invention is used to perform single-block detection on a standard solar cell of a certain size and specification and create a personalized detection standard;
FIG. 5 is an image extracted when the method for detecting geometric appearance defects of solar cells is used for detecting a single solar cell to be detected with a certain size;
FIG. 6 is a detection result image obtained when the method for detecting geometric appearance defects of solar cells is used for detecting a single solar cell to be detected with a certain size and specification;
FIG. 7 is an image extracted when the method for detecting geometric appearance defects of solar cells is used to perform two-block simultaneous detection on standard solar cells of a certain size and specification and create a personalized detection standard;
FIG. 8 is an image extracted when the method for detecting geometric appearance defects of solar cells is used for simultaneously detecting two solar cells to be detected with a certain size specification;
fig. 9 is a detection result image obtained when the geometric appearance defect detection method of the solar cell of the present invention is used to perform two-block simultaneous detection on a solar cell to be detected with a certain size specification.
Detailed Description
In order to make the technical means, the creation features, the achievement purposes and the effects of the invention easy to understand, the following description further explains how the invention is implemented by combining the attached drawings and the detailed implementation modes.
Referring to fig. 1, the method for detecting geometric appearance defects of a solar cell provided by the present invention is divided into two stages, namely a first stage: s1, creating a personalized detection standard; and a second stage: and S2, selecting a personalized detection standard, and detecting the geometrical appearance defects.
S1, creating an individualized detection standard, specifically, creating the individualized detection standard by using the standard solar cell according to the specification and the size of the standard solar cell and the user requirement;
the specification size of the solar cell slice comprises the length and width size and the chamfer size of the solar cell slice; the common length and width dimensions of the standard solar cell sheet include 156mmX156mm, 156mmX78mm, 156mmX52mm, 156mmX26mm and the like, and the common chamfer dimensions include 2mmX2mm, 10mmX10mm and the like;
the user requirement refers to a control standard when a user detects various defects; for example, the user can set the control standard of edge breakage detection to be 1.0mm2When the area of the edge breakage exceeds the set value, the edge breakage is unqualified, and when the area of the edge breakage does not exceed the set value, the edge breakage is qualified;
the standard solar cell is a solar cell with the size meeting the requirement and without unfilled corners, broken edges, holes and crack defects;
the individualized detection standard is set according to the specification and the size of the solar cell and specific requirements of a user, and comprises an allowable range of the length and the width of the solar cell, an allowable range of the size of a chamfer, and judgment values of various defects such as unfilled corners, broken edges, holes, cracks and the like;
s2, selecting a personalized detection standard, and carrying out geometric appearance defect detection: and according to the specification and the size of the solar cell to be detected and the user requirement, selecting an individualized detection standard which is consistent with the model specification and the user detection requirement of the solar cell to be detected from the individualized detection standards created in the step S1.
Referring to fig. 2, in the above-mentioned detection method, step S1 specifically includes the following steps:
s1-1, naming the personalized detection standard;
the naming personalized detection standard specifically refers to that different detection standards are adopted according to different specifications of solar cells and different user requirements, and the detection standards are distinguished through different names;
s1-2, collecting images of the standard solar cell;
the standard solar cell is a solar cell with the size meeting the requirement and without defects of unfilled corners, broken edges, holes and cracks;
the image is generated by a standard solar cell under the backlight source condition, the area of the standard solar cell in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s1-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the standard solar cell image acquired in the step S1-2 by using algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to the step of performing binarization processing on a standard solar cell image subjected to image preprocessing by using a segmentation threshold, wherein the segmentation threshold is preferably 128;
s1-4, selecting all standard solar cell areas and sorting according to positions;
the standard solar cell area is an area obtained after the image segmentation in the step S1-3, the area obtained after the image segmentation in the step S1-3 contains the standard solar cell area and possibly also contains interference areas such as frames and sundries, and when the area obtained after the image segmentation in the step S1-3 contains the interference areas, the standard solar cell area needs to be extracted according to the characteristics such as the length-width ratio, the rectangular degree and the coordinate range of the solar cell and the interference areas are removed;
the sorting according to the position specifically means that when the region obtained after the image segmentation in the step S1-3 includes a plurality of standard solar cell regions, sorting is required according to the respective position, and the position refers to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each standard solar cell region;
s1-5, selecting standard solar cell areas one by one;
specifically, the step of selecting the standard solar cell areas one by one includes the steps of sorting the standard solar cell areas one by one according to positions in the step S1-4, and selecting one standard solar cell area one by one according to the order;
s1-6, inputting the length and width size and the chamfer size of the selected standard solar cell, and setting control parameters such as a defect area threshold, a defect length threshold, a cell length and width size allowable range, a defect number allowable value and the like;
the length and width of the solar cell is specifically the length and width of a rectangle externally connected with the standard solar cell, and the chamfer size is specifically two side lengths of a straight chamfer of the standard solar cell;
wherein, the defect area threshold and the defect length threshold are set according to the specific requirements of users and are used for defect filtration;
when the defect area of the selected standard solar cell to be detected is larger than the set defect area threshold value, determining that the defect area is a defect; otherwise, it is not a defect;
when the maximum diameter of the selected standard solar cell to be detected is larger than a set defect length threshold value, determining that the defect is a defect; otherwise, it is not a defect;
the allowable range of the length and width of the cell slice refers to the allowable range of the length and width of a circumscribed rectangle of a standard solar cell;
when the length and width of the selected standard solar cell slice circumscribed rectangle to be detected exceeds the set range, the standard solar cell slice is an unqualified product;
the defect number allowable value is the maximum value of the sum of the four types of defects, namely corner defect, edge breakage, hole and crack, and the standard solar cell to be detected is qualified when the sum of the four types of defects is smaller than the set value; otherwise, the product is unqualified;
s1-7, extracting four sides of the selected standard solar cell area, calculating the intersection points of the four sides, and obtaining the external quadrangle of the standard solar cell area according to the four intersection points;
the four sides of the standard solar cell area specifically refer to the edge contour of the standard solar cell area extracted by using an edge detection algorithm, and after smooth fitting, the longest line segments on the four sides are respectively selected;
the intersection point of the four sides specifically refers to an intersection point between straight lines where the four longest line segments are located in the edge outline of the extracted standard solar cell area, and a quadrangle determined by the four intersection points is an external quadrangle of the selected standard solar cell area to be obtained;
s1-8, making a difference between the circumscribed quadrangle of the selected standard solar cell area and the selected standard solar cell area, and obtaining four chamfers of the selected standard solar cell;
the four chamfers are obtained by subtracting the selected standard solar cell area from the circumscribed quadrangle of the selected standard solar cell area obtained in the step S1-7, and the chamfer size and the chamfer area of each chamfer are used for judging the corner defect;
s1-9, calculating the length and width scale factor of the circumscribed quadrangle;
the length and width scaling factor is a result obtained by dividing the length and width values of the standard solar cell slice input in the step S1-6 by the long axis and short axis pixel values of the circumscribed quadrangle of the selected standard solar cell slice region obtained in the step S1-7, and the length and width scaling factor establishes a conversion relation between pixels and physical dimensions;
s1-10, calculating the geometric parameters of the selected standard solar cell area by using the length-width scale factor, and storing the geometric parameters into the created personalized detection standard;
the geometric parameters comprise the area, the squareness, the four chamfer sizes and the like, and are specifically the geometric parameters comprising the area, the squareness, the four chamfer sizes and the like obtained after converting pixel units into physical size units through length-width scale factors;
and S1-11, repeating the steps S1-5 to S1-10 until all the standard solar cell areas are processed.
Referring to fig. 3, in the above-mentioned detection method, step S2 specifically includes the following steps:
s2-1, selecting a personalized detection standard;
the individualized detection standard is specifically the individualized detection standard created in step S1, and the selected individualized detection standard is consistent with the model specification of the solar cell to be detected and the user detection requirement;
the method comprises the following steps that control parameters including cell length and width scaling factors, defect area thresholds, defect length thresholds, cell areas, cell chamfer sizes, cell length and width size allowable ranges, defect number allowable values and the like are determined according to personalized detection standards, and whether the solar cell to be detected is qualified or not is judged according to the control parameters;
the length-width scaling factor refers to a conversion relation between a pixel unit and a physical size unit, and is used for converting control parameters such as a defect area, a defect length, a battery piece area, a battery piece chamfer size, a battery piece length-width size and the like from the pixel unit to the physical size unit so as to realize defect judgment;
s2-2, collecting an image of the solar cell to be detected;
the solar cell to be detected is a solar cell which may have geometric appearance defects such as size deviation, unfilled corners, edge breakage, holes or cracks;
the image is an image generated by the solar cell to be detected under the backlight source condition, the area of the solar cell to be detected in the image is close to pure black, the background area is close to pure white, and the image type is a gray image;
s2-3, preprocessing an image and segmenting the image;
the image preprocessing specifically refers to performing image enhancement processing on the solar cell image to be detected acquired in the step S2-2 by adopting algorithms such as denoising, filtering smoothing and contrast enhancement;
the image segmentation specifically refers to performing binarization processing on a solar cell image to be detected after image preprocessing by using a segmentation threshold, wherein the segmentation threshold is preferably 128;
s2-4, selecting all solar cell areas to be detected, and sorting according to positions;
when the region contains the interference region, the solar cell region to be detected needs to be extracted according to the characteristics of the length-width ratio, the rectangular degree, the coordinate range and the like of the cell, and the interference region is removed;
the sorting according to the position means that when the region obtained after the image segmentation in the step S2-3 includes a plurality of solar cell regions to be detected, sorting is required according to the respective position, and the position refers to row and column coordinates of the upper left corner, the upper right corner, the lower left corner, the lower right corner or the center of each solar cell region to be detected;
s2-5, selecting the solar cell areas to be detected one by one;
specifically, the step of selecting the solar cell areas to be detected one by one is to select a certain solar cell area to be detected one by one after sorting according to positions in the step S2-4;
s2-6, calculating the area, the rectangular degree and the length and the width of the minimum circumscribed rectangle of the selected solar cell area to be detected, and converting the area, the rectangular degree and the length and the width into geometric parameters;
the area of the solar cell area to be detected refers to the number of pixels of the area obtained after the binarization processing of the solar cell to be detected; the area value of the normal qualified cell can be changed within a certain range, and the deviation is not too large;
the rectangle degree refers to the similarity degree of the solar cell area to be detected and a rectangle; for a normally qualified battery piece, the rectangle is close to a rectangle, and the rectangle degree is about 1;
the length and width of the minimum external rectangle refers to the length and width of the minimum external rectangle obtained by fitting the solar cell to be detected at any inclination angle; the size of the normal qualified battery piece can be changed within a certain range, and the deviation is not too large;
the geometric parameters are length and width scale factors led in by using the personalized detection standards, and the area, the rectangular degree and the length and width of the minimum external rectangle of the solar cell to be detected are converted into physical size units from pixel units so as to be compared with control parameters in the selected personalized detection standards;
s2-7, performing primary judgment to judge whether the geometric parameters meet the requirements;
if not, determining that the detection is not qualified, and ending the detection;
if so, performing subsequent detection;
the primary determination specifically refers to comparing the geometric parameters obtained in step S2-6 with the control parameters in the personalized test criteria selected in step S2-1, and determining whether the geometric parameters are within an allowable range;
s2-8, extracting four sides of the selected solar cell area to be detected, calculating intersection points of the four sides, and obtaining an external quadrangle of the solar cell area to be detected according to the four intersection points;
the method comprises the following steps of (1) extracting four edges of a selected solar cell area to be detected, wherein the four edges of the selected solar cell area to be detected are subjected to subsequent detection after the initial detection of the step S2-7 is qualified, specifically, an edge contour of the selected solar cell area to be detected is extracted by using an edge detection algorithm, and after smooth fitting, four longest line segments on the edges are respectively selected;
the intersection points of the four edges are intersection points between straight lines where the four longest line segments are located in the edge outline of the selected solar cell area to be detected, and the quadrangle determined by the four intersection points is the external quadrangle of the selected solar cell area to be detected to be obtained;
s2-9, cutting the circumscribed quadrangle of the selected solar cell area to be detected by utilizing the four chamfer sizes and the length and width scale factors in the selected personalized detection standard;
the cutting of the external quadrangle of the solar cell area to be detected specifically means that four chamfers are subtracted from the external quadrangle of the solar cell area to be detected to obtain the solar cell area to be detected without the chamfers, so that the chamfers are prevented from being misjudged as unfilled corners;
s2-10, filling and closing the cut area by using morphology;
the cut area refers to the area of the solar cell to be detected, which is processed in the step S2-9 and does not contain four chamfers;
filling holes and cracks in an area by filling operation, selecting a structural factor with the same direction as that of an external quadrangle, closing a broken edge gap on the edge of the area by closing operation, wherein the external quadrangle after filling and closing operation is a completely convex polygon without four chamfers and any hole cracks and broken edge gaps;
s2-11, making a difference between the morphologically processed region and the selected solar cell region to be detected to obtain a defect candidate region;
the morphologically processed region is specifically the convex polygonal region obtained after the processing in the step S2-10, and the selected to-be-detected cell region is specifically a certain to-be-detected solar cell region selected in the step S2-5; the difference is specifically to subtract the two areas; the defect candidate area is a defect candidate area of all unfilled corners, broken edges, holes and cracks obtained by subtracting the two areas; judging whether the areas are real appearance defects or not, and further judging through S2-12;
s2-12, judging the alternative areas by using the control parameters in the personalized detection standard, and screening out real defects;
the control parameters specifically refer to parameters such as a defect area threshold value, a defect length threshold value and the like which are imported when the personalized detection standard is selected in the step S2-1;
wherein the candidate region is the defect candidate region obtained in step S2-11, which includes all unfilled corners, broken edges, holes and crack defects;
s2-13, classifying the defects according to the positions and the shape characteristic values of the defects;
the defect classification means that the defects are classified into one of unfilled corners, broken edges, holes and cracks respectively according to the positions and the shape characteristic values of the defects; unfilled corner defects are generally located at four chamfered corners; the edge breakage defect is positioned at the edge; the hole defect and the crack defect are both positioned in the battery piece region, the slenderness ratio of the hole defect is smaller, and the slenderness ratio of the crack defect is larger;
s2-14, judging whether the defect number exceeds the limit;
judging whether the defect number exceeds the limit specifically by comparing the detected defect number of the solar cell to be detected with the allowable value of the defect number in the individualized detection standard selected in the step S2-1, judging whether the defect number exceeds the limit, and if the defect number exceeds the limit, judging that the defect number is unqualified; if not, judging the product is qualified;
wherein, the defect number refers to the sum of four types of defects, namely unfilled corners, broken edges, holes and cracks;
s2-15, repeating the steps S2-5 to S2-14 until all the solar cell areas to be detected are processed;
s2-16, storing and outputting the detection result;
the detection result saving specifically refers to saving the detection result, including information such as an actual measurement value, an allowable value range, the number of defects, unqualified reasons and the like, into a database file;
the output specifically refers to sending the detection result to the receiving end in modes such as an MODBUS communication protocol and the like according to a protocol.
Fig. 4 to fig. 6 show a first embodiment of a method for detecting geometric defects of a solar cell according to the present invention;
the embodiment adopts the method for detecting the geometric appearance defects of the solar cell to carry out single-chip detection on the solar cell with a certain size specification;
fig. 4 shows an image region of a standard solar cell obtained in a first stage by using the method for detecting geometric appearance defects of a solar cell provided by the present invention, where the image region includes a first standard
fig. 5 shows an image region of a solar cell to be detected, which is obtained in the second stage by using the method for detecting geometric appearance defects of a solar cell provided by the present invention, where the image region includes a first
FIG. 6 shows a geometric defect inspection result obtained by the geometric defect inspection method for solar cells according to the present invention;
fig. 7 to 8 show a second embodiment of a method for detecting geometric defects of a solar cell according to the present invention;
the embodiment adopts the method for detecting the geometric appearance defects of the solar cell to simultaneously detect the double solar cells with certain dimension and specification;
fig. 7 shows an image region of a standard solar cell obtained in a first stage by using the method for detecting geometric appearance defects of a solar cell provided by the present invention, where the image region includes two second standard
fig. 8 shows an image area of a solar cell to be detected, which is obtained in the second stage by using the method for detecting geometric appearance defects of a solar cell provided by the present invention, where the image area includes two second
Fig. 9 shows a geometric appearance defect detection result obtained by the method for detecting geometric appearance defects of a solar cell provided by the present invention.
Finally, the above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent structures or equivalent processes performed by the present invention or directly or indirectly applied to other related technical fields using the contents of the present specification and the attached drawings are included in the scope of the present invention.