阅读说明：本技术 打印瑕疵检测系统 (Printing flaw detection system ) 是由 李祎 于 2019-10-17 设计创作，主要内容包括：本发明公开了一种打印瑕疵检测系统,包括：照明系统；图像采集系统：将采集的打印产品图像传输给图像识别电脑系统进行识别处理；图像识别电脑系统：对采集的产品图像进行图像识别,根据识别结果判断出打印产品是否有瑕疵,对有瑕疵的打印产品,通过瑕疵报警分级处理装置进行提醒、标记或移出；瑕疵报警分级处理装置：用于当出现有瑕疵的打印产品时进行报警,并对其进行标记或者移出流水线。本发明可以避免传统人工检测导致的漏检、误检以及对检测人员经验的依赖,可以提高出厂产品良率,提高产品质量,降低算法复杂性,提高检测率,降低硬件成本,对于提高产品良率、质量,增加产量提供可靠数据,提高产线的自动化和智能化程度,节约人力成本。(The invention discloses a printing flaw detection system, which comprises: an illumination system; an image acquisition system: transmitting the collected image of the printed product to an image recognition computer system for recognition processing; image recognition computer system: carrying out image recognition on the collected product image, judging whether the printed product has defects according to the recognition result, and reminding, marking or moving out the printed product with defects through a defect alarm grading processing device; flaw alarm grading treatment device: and the printing machine is used for alarming when a defective printed product appears, marking the defective printed product or moving the defective printed product out of the production line. The invention can avoid missing detection, false detection and dependence on detection personnel experience caused by traditional manual detection, can improve the yield of the delivered products, improve the product quality, reduce the algorithm complexity, improve the detection rate and reduce the hardware cost, provides reliable data for improving the yield and quality of the products and increasing the yield, improves the automation and intellectualization degree of a production line and saves the labor cost.)

1. A print defect detection system, comprising:

an illumination system: for providing a light source to the printed product;

an image acquisition system: the lighting system is connected with the image recognition computer system and used for collecting all or part of images of the printed products and transmitting the collected images of the printed products to the image recognition computer system for recognition processing;

image recognition computer system: the image acquisition system is connected with the computer and used for carrying out image identification on the acquired image of the printed product, comparing the image with the image printed by the computer of the printing system, judging whether the printed product has defects according to the identification result, and reminding, marking or moving out the printed product with defects through the defect alarm grading processing device;

flaw alarm grading treatment device: and the image recognition computer system is connected with the image recognition computer system and is used for alarming when defective printed products appear and marking the defective printed products or moving the defective printed products out of a production line.

2. The print defect detection system of claim 1, wherein a plurality of said image capture systems are used to capture images in blocks, and wherein images captured by a plurality of said image capture systems are stitched or processed in blocks.

3. The print defect detection system of claim 2, wherein said image recognition computer system comprises:

a lighting system control module: the lamp is used for controlling the on-off of the lighting lamp;

the image acquisition system reads the module: the image acquisition system is used for controlling the acquisition of images and reading in the images acquired by the image acquisition system;

the flaw detection module: the system is used for detecting whether the acquired image of the printed product has flaws, and if the flaws are found, alarming is carried out through a flaw alarm grading processing device or a flaw marking system is used for marking;

print quality record analysis module: the fault detection device is used for recording and analyzing the fault proportion of a printed product in a specific time period, and alarming when the fault proportion exceeds a set value;

a print quality tracking module: when the system is combined with a printed product identification system, the detection time and the detection result of each printed product are recorded in a database for subsequent inquiry and tracking.

4. The print defect detection system of claim 3, wherein the printer is configured to print separately in several colors, and to illuminate separately with lights of different colors, and to take a picture of each printed single color, and to control the lights of the colors to be turned on and off in sequence.

5. The print defect detection system of claim 3, wherein the images of each color are read individually by controlling a light or the image capture system capturing only images of individual colors.

6. The printing defect detection system of claim 3, wherein said defect comprises at least one of a missing color, a color misalignment, an ink jet, a drop of ink, a string, or a missing line.

7. The print defect detection system of claim 1, wherein said printing system computer sends printed pictures to said image recognition computer system, said image recognition computer system directly comparing said captured printed product image with said printed pictures, reducing computational effort using reduced image size, block detection, region of interest detection, machine learning, or corner detection methods.

8. The print defect detection system of claim 1, wherein for color stitching misalignment or color missing during printing, only the transverse line or position misalignment in the transverse direction is detected, or only the head and tail portion region is detected.

9. The print defect detection system of claim 1, wherein the missing line or the drawn line is detected by a vertical line method for detecting a longitudinal direction.

10. The print defect detection system of claim 1, wherein a smear detection algorithm, a method of reducing image resolution is used for detecting a drop of ink, or a difference in patch area is determined by comparing with an original print image.

Technical Field

The invention relates to the field of printing defect detection, in particular to a printing defect detection system.

Background

Printing technology is currently applied to more and more industrial production fields of wall and floor tiles, cloth, ceramics and the like. Because the inkjet is to spray different colors through a plurality of rows of printing nozzles, different color printing is not printed at the same time or at the same position, patterns with different colors are required to be printed at the same position, but because the position sensor or a part of the color nozzles temporarily fails and the like, the spraying of different colors of the same pattern may have dislocation or some color missing.

Each color nozzle is composed of many fine nozzles or heating blocks, which are easily damaged and may result in large drops of ink on the printed image, or a color missing or streaking.

If the above conditions occur, the defect products are not only caused, but also if the problems are not found in time, a large number of defect products appear later, and the loss is enlarged. At present, a production line using a printing technology is specially provided with a station after an image is printed, and the problems of color deficiency, color dislocation or ink jetting and dripping, wire drawing or line deficiency and the like do not occur in the image through visual detection.

However, the artificial naked eye detection has the following disadvantages: the manual detection is easy to miss detection or generate errors; due to the possibility of problems, training and long-time operation are required to be familiar with the situation, and the product quality and yield depend on the experience of workers; at least 3 manpower is needed for one detection station, and the manpower cost is higher and higher; human eyes are easy to fatigue, and mistakes are easy to make after fatigue.

The printing patterns in the industrial production of wall, floor and cloth materials are complex and special, and a general image recognition algorithm is difficult to find. Some printed patterns imitate marble, granite and the like, and the colors are staggered and complicated. In some patterns, pictures printed by each brick are different, and a printing system can randomly select one picture from a large picture to print, so that the pictures printed by each brick are different. The automatic judgment of the software system is very difficult and complicated, and a general image recognition algorithm is difficult to find.

Disclosure of Invention

The technical problem to be solved by the present invention is to provide a printing defect detection system which can avoid missing detection, false detection and dependence on detection personnel experience caused by traditional manual detection, can improve the yield of the delivered product, improve the product quality, reduce the algorithm complexity, improve the detection rate, reduce the hardware cost, provide reliable data for improving the yield and quality of the product, increase the yield, improve the automation and intelligence degree of the production line, and save the complexity degree of the labor cost, aiming at the above defects of the prior art.

The technical scheme adopted by the invention for solving the technical problems is as follows: a print defect detection system is constructed comprising:

an illumination system: for providing a light source to the printed product;

an image acquisition system: the lighting system is connected with the image recognition computer system and used for collecting all or part of images of the printed products and transmitting the collected images of the printed products to the image recognition computer system for recognition processing;

image recognition computer system: the image acquisition system is connected with the computer and used for carrying out image identification on the acquired image of the printed product, comparing the image with the image printed by the computer of the printing system, judging whether the printed product has defects according to the identification result, and reminding, marking or moving out the printed product with defects through the defect alarm grading processing device;

flaw alarm grading treatment device: and the image recognition computer system is connected with the image recognition computer system and is used for alarming when defective printed products appear and marking the defective printed products or moving the defective printed products out of a production line.

In the printing defect detection system, a plurality of image acquisition systems are adopted to acquire images in a blocking mode, and the images acquired by the image acquisition systems are spliced or subjected to identification blocking processing.

In the print defect detecting system of the present invention, the image recognition computer system comprises:

a lighting system control module: the lamp is used for controlling the on-off of the lighting lamp;

the image acquisition system reads the module: the image acquisition system is used for controlling the acquisition of images and reading in the images acquired by the image acquisition system;

the flaw detection module: the system is used for detecting whether the acquired image of the printed product has flaws, and if the flaws are found, alarming is carried out through a flaw alarm grading processing device or a flaw marking system is used for marking;

print quality record analysis module: the fault detection device is used for recording and analyzing the fault proportion of a printed product in a specific time period, and alarming when the fault proportion exceeds a set value;

a print quality tracking module: when the system is combined with a printed product identification system, the detection time and the detection result of each printed product are recorded in a database for subsequent inquiry and tracking.

In the printing defect detection system, aiming at the condition that a printer is used for printing in several colors separately, lights with different colors are used for irradiating separately, photos with each printing single color are shot, and the lighting lights with each color are controlled to be turned on and off in sequence.

In the printing defect detection system, the light is controlled or the image acquisition system only acquires images of single colors, and the images of all colors are read independently.

In the printing defect detecting system of the present invention, the defect at least includes a color defect, a color dislocation, ink jetting, ink dropping, a string or a line defect.

In the print defect detection system of the present invention, the print system computer sends the printed picture to the image recognition computer system, and the image recognition computer system directly compares the collected image of the printed product with the printed picture, and reduces the amount of calculation by using methods of image size reduction, block detection, key region detection, machine learning, or corner detection.

In the printing defect detection system, aiming at color splicing dislocation or color lack in printing, only transverse lines or position dislocation in the transverse direction is detected, or only the tail part area of the printing head is detected.

In the printing defect detection system, a vertical line detection method in the longitudinal direction is adopted to detect the missing line or the pull line.

In the printing defect detection system, for the ink drop, a stain detection algorithm and a method for reducing the resolution of an image are used for detection, or the difference of small areas is judged by comparing with the original printing image.

The printing defect detection system has the following beneficial effects: the image acquisition system transmits the acquired image of the printed product to the image recognition computer system for recognition processing; the image recognition computer system carries out image recognition on the collected image of the printed product, compares the image with the image printed by the printing system computer, judges whether the printed product has defects according to the recognition result, and reminds, marks or moves out the printed product with defects through a defect alarm grading processing device; the invention can avoid missing detection, false detection and dependence on detection personnel experience caused by traditional manual detection, can improve the yield of the delivered products, improve the product quality, reduce the algorithm complexity, improve the detection rate and reduce the hardware cost, provides reliable data for improving the yield and quality of the products and increasing the yield, improves the automation and intellectualization degree of a production line and saves the labor cost.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic diagram of an embodiment of a print defect detection system according to the present invention;

FIG. 2 is a schematic view of a printing line flow diagram of the printing defect detection system in the embodiment;

FIG. 3 is a schematic structural diagram of an image recognition computer system according to the embodiment.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In an embodiment of the printing defect detecting system of the present invention, a schematic structural diagram of the printing defect detecting system is shown in fig. 1. In FIG. 1, the print defect detection system includes

The lighting system 1 is used for providing a uniform and stable light source for printed products; the misjudgment caused by instability of the collected image due to illumination problems such as uneven light source, shadow and the like is avoided.

The image acquisition system 2 is connected with the lighting system 1 and used for acquiring all or part of images of the printed products and transmitting the acquired images of the printed products to the image recognition computer system 3 for recognition processing. Because some printed products have a large size, such as high requirements on resolution, it may be necessary to acquire images by multiple image acquisition systems 2 in blocks, and to stitch or mark images acquired by multiple image acquisition systems in blocks.

The image recognition computer system 3 is connected with the image acquisition system 2 and is used for carrying out image recognition on the acquired image of the printed product, comparing the image with the image printed by the printing system computer, judging whether the printed product has defects according to the recognition result, and reminding, marking or moving out the printed product with defects through the defect alarm grading processing device 4;

the flaw alarm grading processing device 4 is connected with the image recognition computer system 3 and is used for alarming when a defective printed product appears and marking the printed product or moving the printed product out of a production line.

The printing defect detection system of the invention uses the image acquisition system 2 and the detection software to automatically detect the defects of the printing products and reduce the complexity of the image recognition of the printing defect detection system. Has the following advantages: 1. the missing detection and the false detection caused by the traditional manual detection and the dependence on the experience of detection personnel can be avoided; 2. the yield of the delivered products can be improved, even the quality tracking of each printed product can be provided, and the product quality is improved; 3. aiming at the characteristics of randomness and complexity of a printed image and the appearance of defects of a printed product, image algorithm optimization is carried out, the algorithm complexity is reduced, the detection rate is improved, and the hardware cost is reduced; 4. the method comprises the steps of providing product yield and product proportion of each grade in a specific time period, tracking and analyzing the influence of factors such as production links and production personnel on the product quality, and providing reliable data for improving the product yield and quality and increasing the yield; 5. the automation and the intelligent degree of the production line are improved, and the labor cost is saved.

The invention aims at the defects of manual naked eye detection of the traditional printing production line system, considers the randomness and complexity of the printed image and solves the problem that a general image recognition algorithm is difficult to find. A printing defect detection system is proposed, and fig. 2 is a schematic flow chart of the printing defect detection system applied to a printing production line in the embodiment, and the printing system computer is used for transmitting the printed image to the detection system computer after each printed product is printed. The detection system computer can compare and analyze the printed picture and the actually printed pattern, so that the image algorithm judgment difficulty is reduced.

Fig. 3 is a schematic structural diagram of an image recognition computer system in this embodiment, in fig. 3, the image recognition computer system 3 includes an illumination system control module 31, an image acquisition system reading module 32, a defect detection module 33, a print quality record analysis module 34, and a print quality tracking module 35; wherein the content of the first and second substances,

the lighting system control module 31 is used for controlling the on and off of the lighting light. Because the light attenuation and energy consumption are considered after the light is lightened for a long time, the light can be lightened only when the printed product needs to acquire images, and the on-off of the lighting light needs to be controlled. Aiming at the characteristic that the printer is separately printed in several colors, the light of different colors can be separately irradiated, so that the photos of each printed single color can be shot conveniently, the analysis is convenient, and the lighting lights of the colors need to be controlled to be turned on and off in sequence.

The image acquisition system reading module 32 is responsible for controlling the acquisition of images and reading in images acquired by the image acquisition system 2. Because some printed products have a large size, such as a high requirement on resolution, it may be necessary to acquire images by multiple image acquisition systems 2 in blocks, and at this time, it is necessary to stitch or mark images acquired by multiple image acquisition systems 2 in blocks. Aiming at the characteristic that the printer is separately printed in several colors, only images in single color are collected by controlling light or the image collecting system 2, so that the analysis is convenient, and at the moment, the images in all colors need to be read separately.

The defect detection module 33 is used for detecting whether defects occur in the collected image of the printed product, for example: the flaws include at least color deficiency, color dislocation, ink jetting, ink drop, string or thread deficiency. If a defect is found, the defect is marked by a defect alarm classification processing device 4 alarm or a defect marking system. Because the printed image is random and complex, the size of the printed product is generally large, the acquired image is large, the image recognition algorithm is complex, the calculation amount is large, and in order to simplify the complexity of the image detection algorithm of the printing defect detection system, the following algorithm simplification can be performed according to the characteristics of the printing defect:

1) aiming at the problems of randomness and complexity of the printed image, the printing system computer sends the printed image to the detection system computer, and can directly compare the acquired image with the printed image, wherein the detection calculated amount can be reduced by using methods of reducing the size of the image, or reducing the calculated amount by using methods of blocking detection, key region detection, machine learning, corner point detection and the like.

2) In the printing process, each color is printed in a line and a row respectively, each color needs to be positioned to ensure that the image of each color is printed to the same position, and if a position sensor is in error or the color spraying of individual color fails, color splicing dislocation or color missing occurs in the printing process. So that print color stitching misalignments or color starvation typically occur in the lateral direction and typically occur at the beginning or end of the print. For this characteristic, it is conceivable to detect only a lateral line or a positional misalignment in the lateral direction, or to detect only the head and tail portions of the printhead, thereby reducing the amount of detection calculation.

3) The print head consists of a number of fine nozzles or heated blocks, each dot ejecting a column of colours in the longitudinal direction of the print. If the individual spray head is damaged, a row of colors which are not printed or a row of colors which are directly sprayed can be formed on the printed product, so that the line is lacked and drawn, and a longitudinal vertical line can be seen on the printed product. For this characteristic, it is conceivable to detect a missing line or a pulled line only by detecting a vertical line in the longitudinal direction.

4) In the printing process, part of the printing head is temporarily failed or the ink of the printing head is not cleaned completely, so that color blocks of one point on a printed product can appear, and ink drips can appear. The dripping is generally a stain on a small area, and can be detected by using a stain detection algorithm or even a method for reducing the image resolution. The difference in the small region may be determined by comparison with the original print image.

The printing quality record analysis module 34 is used for recording and analyzing the defect proportion of the printed product in a specific time period, and alarming when the defect proportion exceeds a set value (is higher). The printing quality tracking module 35 is used for recording the detection time and the detection result of each printed product to the database when being combined with the printed product identification system, so that subsequent inquiry tracking can be carried out.

In summary, the present invention automatically detects the defect of the printed product by the print defect detecting system. The image of the printed matter is collected through the image collecting system 2, and the collected image is processed. Aiming at the characteristic of printing by colors, the monochromatic images can be collected by the light control or the image collection device of each color for analysis and processing. And detecting whether the acquired images of the printed products have defects such as color deficiency, color dislocation, ink jetting and dripping, string or line deficiency and the like. Aiming at the problems of randomness and complexity of the printed image, the printing system computer sends the printed image to the detection system computer, and can directly compare and analyze the acquired image and the printed image, thereby simplifying the image detection algorithm and improving the success rate. And correspondingly optimizing and simplifying the image algorithm aiming at the characteristics of each defect. The invention can record and analyze the defect proportion of the printed product in a specific time period, and analyze and improve the conditions of production flow, printer quality, equipment maintenance and the like. When the system is combined with a printed product identification system, the detection time and the detection result of each printed product can be recorded in a database, and subsequent inquiry and tracking can be realized.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.