Material label recognition device

文档序号:1737964 发布日期:2019-12-20 浏览:19次 中文

阅读说明:本技术 物料标签识别装置 (Material label recognition device ) 是由 胡腾飞 于 2019-09-12 设计创作,主要内容包括:本公开揭示了一种物料标签识别装置。其中,物料标签识别装置包括识别机构,为一框形结构,包括框架以及沿所述框架内周缘安装在所述框架内侧的多个标签识别器,所述标签识别器用于识别物料标签;升降机构,与所述识别机构驱动连接;控制机构,分别与所述升降机构以及标签识别器电连接,以分别控制所述升降机构以及标签识别器。这样可以较为方便地识别物料上的标签,特别是对于物料垛上的所有标签,即使所述标签不在一个面上也可以进行识别。(The present disclosure discloses a material label recognition device. The material label identification device comprises an identification mechanism, a label identification device and a label identification device, wherein the identification mechanism is of a frame-shaped structure and comprises a frame and a plurality of label identifiers arranged on the inner side of the frame along the inner peripheral edge of the frame, and the label identifiers are used for identifying material labels; the lifting mechanism is in driving connection with the identification mechanism; and the control mechanism is electrically connected with the lifting mechanism and the label identifier respectively so as to control the lifting mechanism and the label identifier respectively. This allows for easier identification of the labels on the items, particularly for all labels on the stack, even if the labels are not on one side.)

1. A material label identification device, comprising:

the identification mechanism is of a frame-shaped structure and comprises a frame and a plurality of label identifiers arranged on the inner side of the frame along the inner periphery of the frame, and the label identifiers are used for identifying material labels;

the lifting mechanism is in driving connection with the identification mechanism;

and the control mechanism is electrically connected with the lifting mechanism and the label identifier respectively so as to control the lifting mechanism and the label identifier respectively.

2. The material tag identification apparatus of claim 1, wherein the lifting mechanism comprises:

a guide rail;

the sliding block is provided with a guide rail groove, the guide rail penetrates through the guide rail groove and is in sliding connection with the sliding block, and the sliding block is fixedly connected with the identification mechanism;

and the transmission mechanism is in transmission connection with the sliding block and drives the sliding block to move up and down along the guide rail.

3. The material tag identification apparatus of claim 1, wherein the control mechanism comprises:

the lifting mechanism control unit is used for responding to the identification instruction and driving the lifting mechanism to drive the frame to lift;

and the identification mechanism control unit is used for responding to an identification instruction and controlling the label identifier to identify the plurality of material labels on the material.

4. The material tag identification apparatus of claim 3, wherein the identification mechanism control unit comprises:

the image acquisition unit is used for acquiring an image of the material;

the label positioning unit is used for positioning all material labels in the material image;

and the label identification unit is used for identifying all the material labels positioned by the label positioning unit one by one.

5. The material tag identification apparatus of claim 4, wherein the tag positioning unit comprises:

the characteristic vector acquiring unit is used for acquiring a characteristic vector in the material image, wherein the characteristic vector comprises a pixel value and a position coordinate of each pixel point of the material image;

and the machine learning model unit is used for inputting the characteristic vector into a machine learning model, and the machine learning model outputs a set of region position coordinates of each material label on the image.

6. The material label identification device of claim 5, wherein the machine learning model is trained as follows:

acquiring an image sample set comprising a plurality of material labels, wherein each image sample in the image sample set is pasted with a set label containing area position coordinates of all the material labels on the image in advance;

acquiring a feature vector of each image sample in the image sample set;

and inputting the characteristic vector of each image sample in the image sample set into a machine learning model one by one, outputting a set of the region position coordinates of each material label on the image by the machine learning model, judging the set of the region position coordinates of each material label on the image, comparing the set of the region position coordinates of each material label on the image with the attached set label, and if the set of the region position coordinates of each material label on the image output by the machine learning model is inconsistent with the attached set label in advance, adjusting the machine learning model to enable the set of the region position coordinates of each material label on the image output.

7. The material tag identification arrangement of claim 3, wherein the control mechanism further comprises:

the tag quantity acquiring unit is used for acquiring the quantity of the tags of the material to be identified;

the tag quantity confirming unit is used for confirming the quantity of the identified material tags;

and the label quantity judging unit is used for judging whether the quantity of the identified material labels is equal to that of the material labels to be identified or not, and if not, controlling the identifying mechanism to identify the material labels on the material again.

8. The material tag identification arrangement of claim 3, wherein the control mechanism further comprises:

the material position acquisition unit is used for acquiring a material position;

and the material position judging unit is used for judging whether the material is positioned in the area to be identified according to the position of the material, and sending an identification instruction to the lifting mechanism control unit and the identification mechanism control unit if the material is positioned in the area to be identified.

9. The material label identification device of claim 3, further comprising a display module, the control mechanism further comprising:

and the label information display unit is used for controlling and displaying the result of identifying the plurality of material labels on the material or the information borne by the material labels.

10. A material label identification method applied to the material label identification device according to claim 1, comprising:

responding to the identification instruction, and driving the lifting mechanism to drive the frame to lift;

and responding to the identification instruction, and identifying a plurality of material labels on the material.

11. The method for identifying item tags as recited in claim 10, wherein said identifying a plurality of item tags on an item comprises:

acquiring an image of the material;

locating all material tags in the material image;

and identifying all the positioned material labels in the material image one by one.

12. The method for identifying a material label according to claim 11, wherein the positioning of the material label in the material image specifically comprises:

acquiring a characteristic vector in the material image, wherein the characteristic vector comprises a pixel value and a position coordinate of each pixel point of the material image;

and inputting the characteristic vector into a machine learning model, and outputting a set of region position coordinates of each material label on the image by the machine learning model.

13. The method for identifying labels on materials as claimed in claim 12, wherein the machine learning model is trained as follows:

acquiring an image sample set comprising material labels, wherein each image sample in the image sample set is pasted with a set label containing region position coordinates of all the material labels on the image in advance;

acquiring a feature vector of each image sample in the image sample set;

and inputting the characteristic vector of each image sample in the image sample set into a machine learning model one by one, outputting a set of the region position coordinates of each material label on the image by the machine learning model, judging the set of the region position coordinates of each material label on the image, comparing the set of the region position coordinates of each material label on the image with the attached set label, and if the set of the region position coordinates of each material label on the image output by the machine learning model is inconsistent with the attached set label in advance, adjusting the machine learning model to enable the set of the region position coordinates of each material label on the image output.

14. The method of material tag identification as recited in claim 10, the method further comprising:

acquiring the number of the material labels to be identified;

determining a number of the identified material tags;

and judging whether the number of the identified material labels is equal to that of the material labels to be identified or not, and if not, controlling the identification mechanism to identify the material labels on the material again.

15. The material tag identification method of claim 10, wherein prior to driving the elevator mechanism down, the method further comprises:

acquiring the position of a material;

and judging whether the material is positioned in a region to be identified or not according to the position of the material, and if the material is positioned in the region to be scanned, sending an identification instruction so as to drive the lifting mechanism to drive the frame to lift and identify a plurality of material labels on the material.

16. The method of material tag identification as recited in claim 10, the method further comprising:

and controlling a display module to display the result of identifying the plurality of material labels on the material or the information borne by the material labels.

Technical Field

The present disclosure relates to the field of tag identification, and in particular, to a material tag identification apparatus and a material tag identification method.

Background

With the development of the industry and the logistics industry, in the process of material transportation, a label (such as a bar code, a two-dimensional code, an electromagnetic label, or the like) needs to be attached to a material, and the label needs to be identified. At present, labels are often manually scanned and identified one by one, or the labels of materials are arranged in a line or a row in sequence and then are scanned and identified one by a machine according to a preset program, so that the materials need to be manually scanned and identified or the materials need to be manually stacked in the sequence of one line or one row according to the labels.

However, if the material is provided with a plurality of labels, and the plurality of labels are respectively distributed on a plurality of side surfaces, if the material is scanned and identified by using the above method, the labels are not scanned, so that the information of the material labels is wrong, and the accuracy of identifying the information of the material labels is reduced. Therefore, the material label identification needs to be carried out for multiple times, and the efficiency of the material label identification is low and the material label identification is inconvenient.

Disclosure of Invention

In order to solve the problem of low efficiency of identifying material labels which are not on the same plane in the related art, the disclosure provides a material label identification device and a material label identification method.

The present disclosure provides a material label recognition device, the material label recognition device includes:

the identification mechanism is of a frame-shaped structure and comprises a frame and a plurality of label identifiers arranged on the inner side of the frame along the inner periphery of the frame, and the label identifiers are used for identifying material labels;

the lifting mechanism is in driving connection with the identification mechanism;

and the control mechanism is electrically connected with the lifting mechanism and the label identifier respectively so as to control the lifting mechanism and the label identifier respectively.

The present disclosure also provides a material label identification method applied to the material label identification device according to claim 1, including:

driving the lifting mechanism to descend in response to the identification instruction;

and responding to the identification instruction, and identifying the material label on the material.

The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects:

the utility model provides a material label recognition device, material label recognition device includes identification mechanism, elevating system and control mechanism. The identification mechanism is of a frame-shaped structure and comprises a frame and a plurality of label identifiers arranged on the inner side of the frame along the inner peripheral edge of the frame, the label identifiers are used for identifying material labels, and the lifting mechanism is in driving connection with the identification mechanism. And the control mechanism is electrically connected with the lifting mechanism and the label identifier respectively so as to control the lifting mechanism and the label identifier respectively. Therefore, in the process of identifying the material labels, the lifting mechanism is connected with the identification mechanism connected with the lifting mechanism to lift, so that the materials pass through the frame of the identification mechanism, and meanwhile, the plurality of label identifiers of the label identifiers arranged in the frame are arranged around the frame to identify the material labels around the materials, so that the materials can be identified at one time even if the material labels are not on the same plane, and the technical problem of low identification efficiency of the material labels which are not on the same plane is solved.

The material label identification device can conveniently identify the labels on the materials, particularly all the labels on the material stack, even if the labels are not on one surface, the labels can be identified.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

Fig. 1 is a perspective view of one embodiment of the material label identification device.

Fig. 2 is a top view of one embodiment of the material label identification device.

Fig. 3 is a schematic structural view of the control mechanism.

Fig. 4 is a schematic configuration diagram of the identification mechanism control unit.

Fig. 5 is a schematic structural diagram of the label positioning unit.

Fig. 6 is a schematic view of still another structure of the control mechanism.

Fig. 7 is a schematic view of still another structure of the control mechanism.

Fig. 8 is a schematic view of still another structure of the control mechanism.

Fig. 9 is a perspective view of one embodiment of the material label identification device.

Detailed Description

Exemplary embodiments that embody features and advantages of the present disclosure will be described in detail in the following description. It is to be understood that the disclosure is capable of various modifications in various embodiments without departing from the scope of the disclosure, and that the description and drawings are to be regarded as illustrative in nature, and not as restrictive.

The present disclosure provides a material label recognition device, as shown in fig. 1-2, the material label recognition device includes a lifting mechanism 10, a recognition mechanism 20 and a control mechanism 50.

The lifting mechanism 10 is in driving connection with the recognition mechanism 20, and the control mechanism 50 is electrically connected with the lifting mechanism and the recognition mechanism 20 respectively so as to control the lifting mechanism and the recognition mechanism respectively.

The identification mechanism 20 is a frame-shaped structure, and includes a frame 21 and a plurality of tag identifiers 22 mounted on the inner side of the frame 21 along the inner periphery of the frame 21, wherein the tag identifiers 22 are electrically connected to the control mechanism 50 and controlled by the control mechanism 50 for identifying the material tags.

The lifting mechanism 10 includes a guide rail 11, a slider 12, and a transmission mechanism 13. The slider 12 is provided with a guide rail groove 121, the guide rail 11 is inserted into the guide rail groove 121 and is connected with the slider 12 in a sliding manner, and the slider 12 is fixedly connected with one side of the frame 21 of the identification mechanism 20, so that the slider 12 can move together with the identification mechanism 20 when sliding on the guide rail 11. And the transmission mechanism is in transmission connection with the sliding block 12 and drives the sliding block 12 to move up and down along the guide rail.

In one embodiment, as shown in fig. 9, the pair of guide rails 11 is provided, and correspondingly, the pair of guide rail grooves 121 are formed on the slider 12, the guide rails 11 are inserted into the guide rail grooves 121 and slidably connected with the slider 12, one end of each guide rail 11 is fixed on the fixed platform 90, and the other end of each guide rail 11 is far away from the fixed platform 90 and is a free end. The transmission mechanism 13 includes a lifting platform 135, the lifting platform 135 is located on the fixed platform 90 and is fixedly connected with the bottom of the slide block 12, and the lifting platform 135 drives the slide block 12 to lift on the guide rail 11 through self lifting in the lifting process of the lifting mechanism 10.

The fixed platform 90 may be, for example, a flat ground, a platform specially installed, an operation table, a simple support, or the like, as long as the lifting mechanism 10 can be fixed, and the disclosure is not limited thereto.

In another embodiment, as shown in fig. 1-2, the material label identification device further includes a device main body 30, the pair of guide rails 11 are vertically fixed on two vertical sides of the same plane of the device main body 30, correspondingly, the pair of guide rail grooves 121 are formed on the slider 12, and the guide rail 11 is inserted into the guide rail grooves 121 and is slidably connected with the slider 12. The transmission mechanism 13 includes a transmission motor 131, a transmission chain 132 and a fixed pulley 133, the transmission motor 131 and the fixed pulley 133 are respectively located at the upper and lower ends of the same plane of the device main body 30 and are in transmission connection through the transmission chain 132, the transmission chain 132 is fixedly connected with the slider 12, so that when the transmission chain 132 is driven by the rotation motor 131 to move, the transmission chain 132 can drive the slider 12 to lift.

In another embodiment presented in connection with fig. 3, the present disclosure provides that the control mechanism 50 comprises:

a lifting mechanism control unit 510, configured to respond to the identification instruction, and drive the lifting mechanism to drive the frame to lift;

and the identification mechanism control unit 520 is used for responding to the identification instruction and controlling the label identifier to identify the plurality of material labels on the material.

After the material is aligned with the frame 21 of the identification mechanism 20 and then placed, the material identification device can be started to send an identification instruction to the control mechanism 50, and after the lifting mechanism control unit 510 in the control mechanism 50 receives the identification instruction, the lifting mechanism 10 is driven to drive the frame 21 to lift, so that the material passes through the frame 21. Meanwhile, after receiving the identification command, the identification mechanism control unit 520 in the control mechanism 50 controls the tag identifier 22 to identify a plurality of material tags on the material. Thus the tag identifier 22 is identifying a plurality of item tags on the item as it passes through the frame 21.

The method for identifying the plurality of material labels on the material may be that a label identifier 22 is used to obtain material images of the material, and then all the labels in the material images are identified one by one; or may be identified after the label identifier 22 is aligned with the item label.

As shown in fig. 4, in one embodiment of the present disclosure, the identification mechanism control unit 520 includes:

an image acquisition unit 521, configured to acquire an image of the material;

a label positioning unit 522, configured to position all material labels in the material image;

the label identification unit 523 is configured to identify all the material labels located by the label locating unit one by one.

In this embodiment, the method for identifying a plurality of material tags on a material includes acquiring a material image of the material by a tag identifier 22, and then identifying all tags in the material image one by one. The method comprises the specific steps of firstly controlling the tag identifier 22 to acquire an image of the material through the image acquisition unit 521, and then positioning all material tags in the image of the material through the tag positioning unit 522, wherein the material tags can be at different positions on the same surface of the material, so that all the material tags in the image of the material need to be positioned after the image is acquired. After all the material labels in the material image are positioned, all the material labels positioned by the label positioning unit can be identified one by one.

In one embodiment, the specific method for locating all the material labels in the material image may include the following steps: firstly, carrying out gray level processing on the material image, then obtaining gradient amplitudes of the material image in the horizontal and vertical directions, then carrying out average fuzzy processing on the material image according to the gradient amplitudes of the material image in the horizontal and vertical directions, then carrying out image binarization on the material image to obtain a primary binarization image of the material, carrying out multiple expansion and multiple corrosion on the binarization image of the material to obtain a final binarization image of the material, wherein in the final binarization image of the material, an obvious boundary can be displayed in a label area, and finally obtaining the contour position of the material label according to the final binarization image of the material.

As shown in fig. 5, in another embodiment proposed by the present disclosure, the label positioning unit 522 includes:

a feature vector obtaining unit 5221, configured to obtain a feature vector in the material image, where the feature vector includes a pixel value and a position coordinate of each pixel point of the material image;

a machine learning model unit 5222, configured to input the feature vector into a machine learning model, where the machine learning model outputs a set of region position coordinates of each material label on the image.

In this embodiment, all material tags in the material image may be located by a machine learning model, the specific steps can include firstly obtaining a characteristic vector in the material image, wherein the characteristic vector comprises a pixel value and a position coordinate of each pixel point of the material image, a machine learning model can position a coordinate value of each pixel point at the material label boundary in the material image according to the gradient value of each pixel point in the material image and the pixel of the adjacent pixel point, therefore, the regional position coordinates of each material label on the material image can be obtained, and the regional position coordinates can be the position coordinates of the four corners of the material label or a coordinate equation set of the material label boundary. After the regional position coordinates of each material label on the material image are obtained, the machine learning model can output the regional position coordinates of each material label on the material image as a set.

Wherein the machine learning model is trained as follows:

acquiring an image sample set comprising a plurality of material labels, wherein each image sample in the image sample set is pasted with a set label containing area position coordinates of all the material labels on the image in advance;

acquiring a feature vector of each image sample in the image sample set;

and inputting the characteristic vector of each image sample in the image sample set into a machine learning model one by one, outputting a set of the region position coordinates of each material label on the image by the machine learning model, judging the set of the region position coordinates of each material label on the image, comparing the set of the region position coordinates of each material label on the image with the attached set label, and if the set of the region position coordinates of each material label on the image output by the machine learning model is inconsistent with the attached set label in advance, adjusting the machine learning model to enable the set of the region position coordinates of each material label on the image output.

In another embodiment proposed by the present disclosure, as shown in fig. 6, the control mechanism 50 further includes:

a tag number obtaining unit 550, configured to obtain the number of the tags of the material to be identified;

a tag number confirmation unit 560 for determining the number of the identified material tags;

the label quantity judging unit 570 is configured to judge whether the quantity of the identified material labels is equal to the quantity of the material labels to be identified, and if not, control the identifying mechanism to identify the material labels on the material again.

In the process of identifying the material tags, a situation of missing identification may occur, so in this embodiment, after the tag identifier is controlled to identify the plurality of material tags on the material, it is further determined whether the number of identified material tags is equal to the number of material tags to be identified, if not, a situation of erroneous identification or missing identification occurs, and at this time, identification needs to be performed again. In this embodiment, the number of the material tags to be identified is first obtained by the tag number obtaining unit 550, then the number of the identified material tags is determined by the tag number confirming unit 560, and finally, the tag number determining unit 570 determines whether the number of the identified material tags is equal to the number of the material tags to be identified, and if not, the identifying mechanism is controlled to identify the material tags on the material again.

The method for acquiring the number of the material tags to be identified may be a method of manual input by a user, or may be a method of receiving the number of the material tags to be identified sent by an upstream system, which is not limited in this disclosure.

The method for determining the number of the identified material labels can be determined by a counter, namely in the process of identifying the material labels, the counter is added by one in response every time one material label is identified, and after the material labels are identified, the number of the identified material labels can be known by the counter. After the identification of the material labels is completed, the number of the identified material labels can be determined according to the amount of the identified information.

In one embodiment, if the number of the identified material tags is determined by the tag number determining unit 570 to be equal to the number of the material tags to be identified, then it is verified that there is no error identification or missing identification, and then the identification of the material tags can be finished.

In another embodiment proposed by the present disclosure, as shown in fig. 7, the control mechanism 50 further includes:

a material position obtaining unit 580 for obtaining a material position;

the material position judging unit 590 is configured to judge whether the material is located in the area to be identified according to the position of the material, and send an identification instruction to the lifting mechanism control unit and the identification mechanism control unit if the material is located in the area to be identified.

Before carrying out material label discernment, need aim at the material of treating discernment and place the rear behind the frame of identification mechanism and can carry out material label discernment, nevertheless when will treat that the discernment material is placed, can't guarantee each time can both aim at identification mechanism's frame, just at this moment need judge. In this embodiment, a material position is obtained by the material position obtaining unit 580, and then the material position determining unit 590 determines whether the material is located in the area to be identified according to the position of the material, and if the material is located in the area to be identified, an identification instruction is sent to the lifting mechanism control unit and the identification mechanism control unit, so that the lifting mechanism control unit 510 drives the lifting mechanism 10 to drive the frame 21 to lift in response to the identification instruction. And the identification mechanism control unit 520 controls the tag identifier 22 to identify the plurality of material tags on the material in response to the identification instruction.

The method for obtaining the position of the material may be to add a locator to the material, determine the position of the material according to the locator, obtain a position image of the material, determine the position of the material according to a real-time image of the material, and determine the position of the material by determining the distance of the material from a distance sensor, which is not limited in this disclosure. In this embodiment, a plurality of distance sensors are disposed in the material tag identification device, and the distance between the material tag identification device and the distance sensors is obtained by the distance sensors, so as to determine the position of the material.

Referring to fig. 8, in another embodiment of the present disclosure, the material label identification apparatus further includes a display module 80, and the control mechanism 50 further includes:

and a tag information display unit 540, configured to control to display a result of identifying the material tag or information carried by the material tag.

In this embodiment, the material label identification apparatus further includes a display module 80, so that the control mechanism 50 can control the label information display unit 540 to display the identification result of the material label or the information carried by the material label through the display module 80. The result of identifying the material tags may include the number of the material tags identified by the material tag identification, the number of the material tags still to be identified by the material tag or whether the material tag identification is successful, and the number and the position information of the material tags not identified in the material tag identification process. Therefore, related operators can adjust the material or the material label identification device according to the identification result of the material label displayed by the display module 80, and the material label identification device can be normally used.

The present disclosure further provides a material tag identification method, including:

responding to the identification instruction, and driving the lifting mechanism to drive the frame to lift;

and responding to the identification instruction, and identifying a plurality of material labels on the material.

After the material is aligned with the frame 21 of the identification mechanism 20 and then placed, the material identification device can be started to send an identification instruction to the control mechanism 50, and after the control mechanism 50 receives the identification instruction, the lifting mechanism 10 is driven to drive the frame 21 to lift, so that the material passes through the frame 21. Meanwhile, the tag identifier 22 is controlled to identify a plurality of material tags on the material. Thus the tag identifier 22 is identifying a plurality of item tags on the item as it passes through the frame 21.

The method for identifying the plurality of material labels on the material may be that a label identifier 22 is used to obtain material images of the material, and then all the labels in the material images are identified one by one; or may be identified after the label identifier 22 is aligned with the item label.

In one embodiment, the specific step of identifying the plurality of material tags on the material includes:

acquiring an image of the material;

locating all material tags in the material image;

and identifying all the positioned material labels in the material image one by one.

In this embodiment, the method for identifying a plurality of material tags on a material includes acquiring a material image of the material by a tag identifier 22, and then identifying all tags in the material image one by one. The specific steps of the method include that the label recognizer 22 is controlled to obtain an image of the material, and then all material labels in the image of the material are located, the material labels may be at different positions on the same surface of the material, so that all the material labels in the image of the material need to be located after the image is obtained, the location method may be to locate position coordinates of four corners of each material label, or may be to locate a coordinate equation of a boundary line of the material labels, the position coordinates may be cartesian coordinates or polar coordinates, and the disclosure is not limited herein. After all the material labels in the material image are positioned, all the material labels positioned by the label positioning unit can be identified one by one.

In one embodiment, the specific method for locating all the material labels in the material image may include the following steps: firstly, carrying out gray level processing on the material image, then obtaining gradient amplitudes of the material image in the horizontal and vertical directions, then carrying out average fuzzy processing on the material image according to the gradient amplitudes of the material image in the horizontal and vertical directions, then carrying out image binarization on the material image to obtain a primary binarization image of the material, carrying out multiple expansion and multiple corrosion on the binarization image of the material to obtain a final binarization image of the material, wherein in the final binarization image of the material, an obvious boundary can be displayed in a label area, and finally obtaining the contour position of the material label according to the final binarization image of the material.

In one embodiment, the specific step of locating the material label in the material image includes:

acquiring a characteristic vector in the material image, wherein the characteristic vector comprises a pixel value and a position coordinate of each pixel point of the material image;

and inputting the characteristic vector into a machine learning model, and outputting a set of region position coordinates of each material label on the image by the machine learning model.

In this embodiment, all material tags in the material image may be located by a machine learning model, the specific steps can include firstly obtaining a characteristic vector in the material image, wherein the characteristic vector comprises a pixel value and a position coordinate of each pixel point of the material image, a machine learning model can position a coordinate value of each pixel point at the material label boundary in the material image according to the gradient value of each pixel point in the material image and the pixel of the adjacent pixel point, therefore, the regional position coordinates of each material label on the material image can be obtained, and the regional position coordinates can be the position coordinates of the four corners of the material label or a coordinate equation set of the material label boundary. After the regional position coordinates of each material label on the material image are obtained, the machine learning model can output the regional position coordinates of each material label on the material image as a set.

The machine learning model is trained as follows:

acquiring an image sample set comprising material labels, wherein each image sample in the image sample set is pasted with a set label containing region position coordinates of all the material labels on the image in advance;

acquiring a feature vector of each image sample in the image sample set;

and inputting the characteristic vector of each image sample in the image sample set into a machine learning model one by one, outputting a set of the region position coordinates of each material label on the image by the machine learning model, judging the set of the region position coordinates of each material label on the image, comparing the set of the region position coordinates of each material label on the image with the attached set label, and if the set of the region position coordinates of each material label on the image output by the machine learning model is inconsistent with the attached set label in advance, adjusting the machine learning model to enable the set of the region position coordinates of each material label on the image output.

In one embodiment, the method further comprises:

acquiring the number of the material labels to be identified;

determining a number of the identified material tags;

and judging whether the number of the identified material labels is equal to that of the material labels to be identified or not, and if not, controlling the identification mechanism to identify the material labels on the material again.

In the process of identifying the material tags, a situation of missing identification may occur, so in this embodiment, after the tag identifier is controlled to identify the plurality of material tags on the material, it is further determined whether the number of identified material tags is equal to the number of material tags to be identified, if not, a situation of erroneous identification or missing identification occurs, and at this time, identification needs to be performed again. In this embodiment, the number of the material tags to be identified is specifically obtained first, then the number of the identified material tags is determined, and finally, whether the number of the identified material tags is equal to the number of the material tags to be identified is judged, and if not, the identifying mechanism is controlled to identify the material tags on the material again. The method for acquiring the number of the material tags to be identified may be a method of manual input by a user, or may be a method of receiving the number of the material tags to be identified sent by an upstream system, which is not limited in this disclosure.

The method for determining the number of the identified material labels can be determined by a counter, namely in the process of identifying the material labels, the counter is added by one in response every time one material label is identified, and after the material labels are identified, the number of the identified material labels can be known by the counter. After the identification of the material labels is completed, the number of the identified material labels can be determined according to the amount of the identified information.

In one embodiment, if the number of the identified material tags is determined by the tag number determining unit 570 to be equal to the number of the material tags to be identified, then it is verified that there is no error identification or missing identification, and then the identification of the material tags can be finished.

In one embodiment, before driving the lifting mechanism to descend, the method further comprises:

acquiring the position of a material;

and judging whether the material is positioned in a region to be identified or not according to the position of the material, and if the material is positioned in the region to be scanned, sending an identification instruction so as to drive the lifting mechanism to drive the frame to lift and identify a plurality of material labels on the material.

In one embodiment, the method further comprises:

and controlling a display module to display the scanning result and the information borne by the material label.

In this embodiment, the material label recognition device further includes a display module 80, and the result of the material label recognition or the information carried by the material label can be displayed through the display module 80. The result of identifying the material tags may include the number of the material tags identified by the material tag identification, the number of the material tags still to be identified by the material tag or whether the material tag identification is successful, and the number and the position information of the material tags not identified in the material tag identification process. Therefore, related operators can adjust the material or the material label identification device according to the identification result of the material label displayed by the display module 80, and the material label identification device can be normally used.

While the present disclosure has been described with reference to several exemplary embodiments, it is understood that the terminology used is intended to be in the nature of words of description and illustration, rather than of limitation. As the present disclosure may be embodied in several forms without departing from the spirit or essential characteristics thereof, it should also be understood that the above-described embodiments are not limited by any of the details of the foregoing description, but rather should be construed broadly within its spirit and scope as defined in the appended claims, and therefore all changes and modifications that fall within the meets and bounds of the claims, or equivalences of such meets and bounds are therefore intended to be embraced by the appended claims.

16页详细技术资料下载
上一篇:一种医用注射器针头装配设备
下一篇:对近场通信NFC设备进行相位校准的方法、设备和系统

网友询问留言

已有0条留言

还没有人留言评论。精彩留言会获得点赞!

精彩留言,会给你点赞!