Non-grinding rework processing method for circuit redness defective plate

By using visual modules and image analysis-driven differential processing, the problem of insufficient ink thickness caused by the reddish color of the circuit board was solved, achieving grinding-free processing, improving yield and ink adhesion, and simplifying the process flow.

CN121940969APending Publication Date: 2026-04-28ZHUHAI ZHENWANG ELECTRONICS CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
ZHUHAI ZHENWANG ELECTRONICS CO LTD
Filing Date
2025-12-31
Publication Date
2026-04-28

AI Technical Summary

Technical Problem

In traditional processes, the reddish defect on the circuit board leads to insufficient ink thickness, resulting in complex subsequent processing and a high risk of oxidation and corrosion of the finished product. Existing methods usually require grinding and rework, which increases the complexity and cost of the process.

Method used

A vision module is used to acquire circuit board images. Features are determined through image analysis, and differentiated brushing, pickling, and sandblasting processes are implemented to avoid grinding. Specialized treatments are applied to reddened areas to ensure consistent ink thickness.

Benefits of technology

This technology enables the non-grinding process of red-tinted circuit boards, improving the finished product qualification rate, reducing costs, simplifying the processing flow, and enhancing the ink adhesion and consistency of the finished products.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121940969A_ABST
    Figure CN121940969A_ABST
Patent Text Reader

Abstract

The invention discloses a non-grinding rework processing method for a circuit redness defect board, and relates to the field of semiconductor processing, the non-grinding rework processing method for the circuit redness defect board is characterized in that a redness circuit board and a normal circuit board are processed on the same production line by upgrading and improving the traditional circuit board polish-brush, acid etching and sand blasting processing flow, and the circuit board with the circuit redness defect is processed on the same production line. In the whole scheme, the visual scheme is technically driven, an image analysis method is adopted, ink characteristics of the circuit board are determined, and then according to a preset scheme, differentiated processing flows are adopted for different characteristic conditions of different circuit boards, so that collinear processing of all the circuit boards can be finally realized; a grinding-free and reworking-free synchronous processing flow is carried out on the redness circuit board, so that the qualification rate of the whole circuit board is improved, the damage to goods is reduced, meanwhile, a complex processing production line does not need to be additionally arranged for the redness circuit board, and the cost is effectively saved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of semiconductor processing, and in particular to a non-grinding rework method for circuit board defects caused by reddening. Background Technology

[0002] During the manufacturing process of PCB circuit boards, solder resist ink is applied to the surface of the resin glass fiber layer. The core function of the solder resist ink is to isolate the resin glass fiber from external corrosive factors such as heat, oxygen, and moisture. The thickness of the ink directly determines the strength of the protective ability. When the ink thickness meets the standard (typical dry film thickness 20-35um), it can effectively block the conduction of soldering heat to the substrate, while isolating oxygen and water vapor to prevent the resin glass fiber of the substrate from oxidizing and decomposing. When the ink becomes thin (<15um), or even when there are local omissions or pinholes, heat, oxygen, and water vapor can easily penetrate the ink layer and act directly on the substrate resin. After the resin undergoes thermal or chemical oxidation, it will gradually carbonize and change color, appearing red. The thinner the ink, the more fully the corrosive factors come into contact with the substrate, the more intense the oxidation reaction, and the deeper and larger the red color.

[0003] In traditional processes, circuit board manufacturing involves grinding, acid etching, and sandblasting to remove surface oxide layers, impurities, and irregular ink. This improves adhesion for subsequent touch-up ink coating. However, for reddish circuit boards, the ink thickness is significantly lower than the standard thickness. Following a standardized process would damage the base ink layer and even harm the circuit board. Without treatment, subsequent ink adhesion is difficult, increasing the risk of oxidation and corrosion in the finished product. Therefore, current processes typically require rework before and after grinding to increase the thickness of the base ink layer, resulting in a standardized substrate after subsequent acid washing and sandblasting. This makes the entire process cumbersome, requiring multiple processing and inspection steps, leading many manufacturers to discard the boards. To address this issue, a non-grinding rework method for reddish circuit board defects is provided. Summary of the Invention

[0004] The purpose of this invention is to provide a non-grinding rework method for circuit board with red-hot defects, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for non-grinding rework of a circuit board with a reddish tint, comprising the following steps: S1. Loading: The vision module acquires images of the circuit board, and the images are transmitted to the analysis module for image analysis. S2. Extract relevant features based on image analysis, and generate control signals for brushing, pickling and sandblasting treatments based on the features respectively; S3. The circuit board enters the grinding, pickling and sandblasting processes in sequence, and the corresponding operation is performed according to the analysis results. The red-colored circuit board uses a non-grinding process; Before entering the pickling process, the area of ​​the reddened area on the circuit board is judged to determine whether to enter the spray pickling or immersion pickling tank for pickling treatment. When entering the sandblasting process, special targeted sandblasting treatment is adopted for the red-hot circuit boards; S4. Before unloading, the visual module is used to detect whether the material is qualified, and then the material is unloaded according to its classification.

[0006] Preferably, in step S1, the circuit board is continuously fed through a uniform feeding conveyor belt. A vision module is installed across the end of the feeding conveyor belt. The vision module acquires images of the surface of the circuit board passing below it. An industrial computer is designed to connect with the vision module for signal transmission. Professional image analysis software is installed in the industrial computer. The acquired images are then transmitted to the industrial computer, and image features are extracted by quickly analyzing the acquired images.

[0007] Preferably, in step S2, the features extracted after image analysis include the presence of redness, the ratio of the red area to the total area, the depth of the redness, and abnormal color spot features. These features are analyzed and judged using image analysis software. During image processing, image segmentation is used to divide the image into multiple rectangular regions, and then a grid method is used to divide the smallest image acquisition unit. During color judgment, multiple intervals are divided by setting a color threshold, thereby judging the degree of oxidation of the resin glass fiber based on color. By extracting abnormal color spots, it is determined that there is an impurity accumulation problem based on the color spots.

[0008] Preferably, the image analysis software uses a preset database and data model to pre-store images of various conditions existing on the circuit board in the database for training the data model, so as to better extract fixed features from the images, and determine the specific working condition of the circuit board based on the features, and then take targeted special processing measures.

[0009] Preferably, the operation actions are controlled based on image features to control the brushing, pickling and sandblasting processes. During image analysis, it is determined whether there is redness by each smallest unit. When any smallest unit is red, it is determined to be a "red circuit board", and a brushless solution is adopted in the brushing chamber. When the number of reddened smallest units reaches 95% of the total number of smallest units, it is judged as "weak homogenization" and "short-time immersion pickling" is adopted in the pickling tank. If it is less than 95%, "point-to-point spray pickling" is adopted. When concentrated spots are present in certain areas, and the number of the smallest units containing spots in the concentrated area reaches the minimum spot area threshold, the area is marked as an impurity area. During the sandblasting process, this area is subjected to additional concentrated sandblasting cleaning.

[0010] Preferably, the brush chamber is equipped with a liftable brush component, which allows the circuit board to pass through the brush chamber; If it is determined to be a "red-light circuit board", the brush will rise to its highest point. At this time, the brush will not be in contact with the "red-light circuit board" and the brush will not need to be driven. If the circuit board is determined to be "non-red-emitting circuit board", the ink thickness is determined based on the strong correlation between the thickness of the solder resist ink and its surface gloss, grayscale value, and texture uniformity. The image analysis software removes dust and noise through image filtering, converts the color image to a grayscale image through grayscale processing, and extracts the ink area and background through binarization segmentation. Then, it calculates the average grayscale value of all pixels in the detection area (value 0-255); extracts the standard deviation of gloss in the image; identifies the proportion of the transparent area of ​​the glass fiber texture on the board surface, and then performs brush control, including... Grayscale value 180-220+ standard deviation <10+ transparency ratio <3%, ink thickness is judged to be 20-35μm, surface is flat, no texture is visible, ink thickness is standard thickness, at this time the brush pressure is 0.35MPa; brush wheel speed is 900r / min; board surface conveying speed is 1.5m / min; brushing time is 14s; Grayscale value 220-230+, standard deviation 10-20+, transparency ratio 3%-10%, ink thickness judged to be 15-20um, slight transparency in some areas, no reddish tinge, ink thickness is relatively thin, at this time the brush pressure is 0.175MPa; brush wheel speed 700r / min; board surface conveying speed 2m / min; brushing time 9s; Grayscale value 230-240+ standard deviation >20+ transparency ratio >10%, indicating ink thickness of 10-15um, large area texture transparency, ink thickness close to the critical value, at this time the brush pressure is 0.13MPa; brush wheel speed is 500r / min; board surface conveying speed is 2.5m / min; brushing time is 5.5s.

[0011] Preferably, in step S3, a diversion scheme is adopted to transport the circuit board to the spray pickling tank and the immersion pickling tank respectively through two conveyor belts. A driving component is designed at the diversion point, and a control signal for the driving component is generated based on the image feature analysis results so that the driving component pushes the circuit board into the spray pickling tank or the immersion pickling tank. Since both red-tinted and non-red-tinted circuit boards will be immersed in the same pickling tank for pickling, the pickling tank is set with a sulfuric acid concentration of 7%-9% and a temperature of 25-30℃. For red-tinted circuit boards that are judged to have "weak homogenization", the immersion time in the pickling tank is ≤25s; for non-red-tinted circuit boards, the immersion time in the pickling tank is ≥45s.

[0012] Preferably, for the reddened circuit boards entering the spray pickling tank, a differentiated spraying scheme is adopted based on the non-reddened area, reddened area, and different chromaticities of the reddened area determined by image segmentation, including: In non-reddening areas, a spray pressure of 0.3 MPa is used for a spraying time of 25 seconds; For areas with slight redness, spray at a pressure of 0.25 MPa for 25 seconds. For areas with moderate reddening, use a spray pressure of 0.15 MPa and a spray time of 15 seconds. For areas with severe redness, use a spray pressure of 0.12 MPa and a spray time of 15 seconds. By using multi-point nozzles for the spray area, all nozzles are supplied with high-pressure acid via a plunger pump, and pressure and spraying time are controlled by connecting precision pressure reducing valves and stop valves respectively.

[0013] Preferably, when the sandblasting chamber performs sandblasting on circuit boards, for circuit boards without red glow, a differentiated spraying scheme is adopted based on the ink thickness, wherein the sandblasting pressure and sandblasting time decrease as the ink thickness decreases; For red-tinted circuit boards, after pre-treatment of impurity areas using a reinforced sandblasting method, a uniform rapid low-pressure sandblasting is performed to ensure impurity removal while improving the consistency of polishing the red-tinted circuit boards.

[0014] Preferably, in step S4, another set of detection modules is set on the unloading conveyor belt at the outlet of the sandblasting chamber. Before unloading, the circuit board image is acquired by the vision detection module, and then the image analysis software performs image analysis to make a secondary judgment on the ink thickness of the circuit board. If the ink thickness is lower than the minimum threshold standard, it is judged as waste material; otherwise, it is judged as qualified material, and then it is sorted and unloaded.

[0015] The technical effects and advantages of this invention are as follows: 1. This method for reworking red-colored defective circuit boards without grinding upgrades and improves upon traditional circuit board grinding, acid etching, and sandblasting processes. Red-colored circuit boards are processed on the same production line as normal circuit boards. The entire solution is technology-driven, using image analysis to determine the ink characteristics of the circuit boards. Then, based on a pre-set plan, differentiated processing procedures are adopted for different characteristics of different circuit boards, ensuring that all circuit boards can be processed on the same line. Red-colored circuit boards undergo simultaneous processing without grinding or rework, thereby improving the overall circuit board pass rate, reducing material loss, and eliminating the need for additional complex processing lines for red-colored circuit boards, effectively saving costs.

[0016] 2. The non-grinding rework method for the red-colored defective circuit board is as follows: For the red-colored circuit board, after image analysis, non-grinding treatment is carried out in the brush chamber. In the pickling area, the spraying or immersion pickling scheme is determined according to the area ratio of the red area, and different spraying schemes are adopted according to the degree of redness to avoid over-etching of weak areas. When entering the sandblasting area, the impurity area is pre-treated with a sandblasting scheme, and then uniformly subjected to rapid low-pressure sandblasting. This makes the finished ink layer achieve the effect of removing impurities and surface brushing to improve the adhesion of subsequent inks. At the same time, the original uneven ink layer achieves the effect of uniform thickness through targeted treatment, thereby improving the yield of finished products.

[0017] 3. The non-grinding rework method for circuit boards with reddish defects involves using image processing technology for image recognition. Simultaneously, based on the strong correlation between the thickness of the solder resist ink and its surface gloss, grayscale value, and texture uniformity, the ink thickness is determined. The ink thickness is judged according to the grayscale value, gloss standard deviation, and transparency ratio. A processing intensity strategy positively correlated with ink thickness is adopted during brushing and sandblasting to further improve the consistency of the finished product. Attached Figure Description

[0018] Figure 1 This is a flowchart of the overall process of the non-grinding rework method of the present invention; Figure 2 This is a flowchart of the analysis and control logic of the detection module of the present invention; Figure 3 This is a simplified schematic diagram of the overall operation process of the non-grinding rework method of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention provides, for example Figures 1-3 The method shown is a non-grinding rework process for a circuit board with a reddish tint defect, comprising the following steps: S1. Loading: The vision module acquires images of the circuit board, and the images are transmitted to the analysis module for image analysis. S2. Extract relevant features based on image analysis, and generate control signals for brushing, pickling and sandblasting treatments based on the features respectively; S3. The circuit board enters the grinding, pickling and sandblasting processes in sequence, and the corresponding operation is performed according to the analysis results. The red-colored circuit board uses a non-grinding process; Before entering the pickling process, the area of ​​the reddened area on the circuit board is judged to determine whether to enter the spray pickling or immersion pickling tank for pickling treatment. When entering the sandblasting process, special targeted sandblasting treatment is adopted for the red-hot circuit boards; S4. Before unloading, the visual module is used to detect whether the material is qualified, and then the material is unloaded according to its classification.

[0021] In step S1, the circuit boards are continuously fed through a uniform feeding conveyor belt. At the end of the feeding conveyor belt, a vision module is set across it. The vision module acquires images of the surface of the circuit boards passing below it. The vision module is connected to an industrial computer for signal transmission. Professional image analysis software is installed in the industrial computer, and the acquired images are then transmitted to the industrial computer. The acquired images are quickly analyzed to extract image features.

[0022] In step S2, the features extracted after image analysis include the presence of redness, the ratio of the red area to the total area, the depth of the redness, and abnormal color spot features. These features are analyzed and judged using image analysis software. During image processing, image segmentation is used to divide the image into multiple rectangular regions, and then a grid method is used to divide the smallest image acquisition unit. In the color judgment process, multiple intervals are divided by setting a color threshold, thereby judging the degree of oxidation of the resin glass fiber based on the color. By extracting abnormal color spots, it is determined that there is an impurity accumulation problem based on the color spots.

[0023] Image analysis software uses a pre-set database and data model to store images of various circuit board conditions in the database for training the data model. This allows for better extraction of fixed features from the images, and the software can then determine the specific operating condition of the circuit board based on these features, followed by targeted processing measures.

[0024] The operation is controlled based on image features to control the brushing, pickling and sandblasting processes. During image analysis, it is determined whether there is redness by each smallest unit. When any smallest unit is red, it is judged as a "red circuit board". At this time, a brushless solution is adopted in the brush chamber. When the number of reddened smallest units reaches 95% of the total number of smallest units, it is judged as "weak homogenization" and "short-time immersion pickling" is adopted in the pickling tank. If it is less than 95%, "point-to-point spray pickling" is adopted. When concentrated spots are present in certain areas, and the number of the smallest units containing spots in the concentrated area reaches the minimum spot area threshold, the area is marked as an impurity area. During the sandblasting process, this area is subjected to additional concentrated sandblasting cleaning.

[0025] The brush chamber is equipped with liftable brushes, which move as the circuit board passes through the brush chamber. If it is determined to be a "red-light circuit board", the brush will rise to its highest point. At this time, the brush will not be in contact with the "red-light circuit board" and the brush will not need to be driven. If the circuit board is determined to be "non-red-emitting circuit board", the ink thickness is determined based on the strong correlation between the thickness of the solder resist ink and its surface gloss, grayscale value, and texture uniformity. The image analysis software removes dust and noise through image filtering, converts the color image to a grayscale image through grayscale processing, and extracts the ink area and background through binarization segmentation. Then, it calculates the average grayscale value of all pixels in the detection area (value 0-255); extracts the standard deviation of gloss in the image; identifies the proportion of the transparent area of ​​the glass fiber texture on the board surface, and then performs brush control, including... Grayscale value 180-220+ standard deviation <10+ transparency ratio <3%, ink thickness is judged to be 20-35μm, surface is flat, no texture is visible, ink thickness is standard thickness, at this time the brush pressure is 0.35MPa; brush wheel speed is 900r / min; board surface conveying speed is 1.5m / min; brushing time is 14s; Grayscale value 220-230+, standard deviation 10-20+, transparency ratio 3%-10%, ink thickness judged to be 15-20um, slight transparency in some areas, no reddish tinge, ink thickness is relatively thin, at this time the brush pressure is 0.175MPa; brush wheel speed 700r / min; board surface conveying speed 2m / min; brushing time 9s; Grayscale value 230-240+ standard deviation >20+ transparency ratio >10%, indicating ink thickness of 10-15um, large area texture transparency, ink thickness close to the critical value, at this time the brush pressure is 0.13MPa; brush wheel speed is 500r / min; board surface conveying speed is 2.5m / min; brushing time is 5.5s.

[0026] During the grayscale measurement process, the corresponding grayscale values ​​are obtained by taking pictures with a vision camera and then processing them. The standard deviation of the photometric measurement is calculated by measuring the physical gloss of the ink layer in different areas of the circuit board with a standard gloss meter, and the unit is GU. In step S3, a diversion scheme is adopted to transport the circuit board to the spray pickling tank and the immersion pickling tank through two conveyor belts respectively. A driving component is designed at the diversion point. The driving component control signal is generated based on the image feature analysis results so that the driving component can push the circuit board into the spray pickling tank or the immersion pickling tank. Since both red-tinted and non-red-tinted circuit boards will be immersed in the same pickling tank for pickling, the pickling tank is set with a sulfuric acid concentration of 7%-9% and a temperature of 25-30℃. For red-tinted circuit boards that are judged to have "weak homogenization", the immersion time in the pickling tank is ≤25s; for non-red-tinted circuit boards, the immersion time in the pickling tank is ≥45s.

[0027] For the reddened circuit boards entering the spray pickling tank, a differentiated spraying scheme is adopted based on the non-reddened area, reddened area, and different chromaticities of the reddened area determined by image segmentation, including: In non-reddening areas, a spray pressure of 0.3 MPa is used for a spraying time of 25 seconds; For areas with slight redness, spray at a pressure of 0.25 MPa for 25 seconds. For areas with moderate reddening, use a spray pressure of 0.15 MPa and a spray time of 15 seconds. For areas with severe redness, use a spray pressure of 0.12 MPa and a spray time of 15 seconds. By using multi-point nozzles for the spray area, all nozzles are supplied with high-pressure acid via a plunger pump, and pressure and spraying time are controlled by connecting precision pressure reducing valves and stop valves respectively.

[0028] When sandblasting circuit boards, the sandblasting chamber adopts a differentiated spraying scheme based on the ink thickness for circuit boards without red glow. The sandblasting pressure and sandblasting time decrease as the ink thickness decreases. For red-colored circuit boards, after pre-treatment of impurity areas using a reinforced sandblasting scheme, a uniform, rapid, low-pressure sandblasting is performed. This ensures impurity removal while improving the consistency of the red-colored circuit board polishing. By controlling the angle and sandblasting pressure of the spray nozzle, differentiated spraying cleaning is achieved for different impurity areas to remove impurities. Then, uniform sandblasting is performed on the entire board to achieve a uniform and consistent overall surface.

[0029] In step S4, another set of detection modules is set on the unloading conveyor belt at the outlet of the sandblasting chamber. Before unloading, the circuit board image is acquired by the vision detection module, and then the image analysis software performs image analysis to make a secondary judgment on the ink thickness of the circuit board. If the ink thickness is lower than the minimum threshold standard, it is judged as waste material; otherwise, it is judged as qualified material. Then, the materials are sorted and unloaded. By setting a threshold for the final ink layer thickness of the red-colored circuit board and the non-red-colored circuit board, the quality is judged based on the detection results. For qualified red-colored circuit boards and non-red-colored circuit boards, a differentiated secondary ink spraying scheme is finally adopted. The red-colored circuit board is over-compensated to make the final product more consistent.

[0030] Working principle: This method for non-grinding rework of circuit board reddening defects is an upgrade and improvement of the traditional circuit board grinding, acid etching and sandblasting process. The reddening circuit board is reworked on the same production line as the normal circuit board without grinding. By modifying and upgrading the production line, during the material feeding stage, the vision module acquires the image information of each circuit board. Based on the image information, relevant features are extracted, including whether reddening exists, the ratio of the reddened area to the total area, the depth of reddening, and abnormal color spot features. Based on the features, control signals for grinding, acid etching and sandblasting are generated respectively. When the circuit board enters the brush chamber, the "whether the circuit board is red" characteristic is used as the judgment criterion. When the circuit board is red, the brush rises to the highest point. At this time, the brush does not contact the "red circuit board" and the brush does not need to be driven. If the circuit board is determined to be "non-red-emitting," the ink thickness is judged based on the strong correlation between the thickness of the solder resist ink and its surface gloss, grayscale value, and texture uniformity. When the grayscale value is 180-220 + standard deviation < 10 + transparency < 3%, the ink thickness is judged to be 20-35 μm. The surface is smooth, with no texture transparency, and the ink thickness is considered standard. In this case, the brush pressure is 0.35 MPa; the brush wheel speed is 900 r / min; the board surface conveying speed is 1.5 m / min; and the brushing time is 14 s. When the grayscale value is 220-230 + standard deviation 10-20 + transparency 3%-10%, the ink thickness is judged to be 15-2... 0µm, slight local transparency, no reddish tint, ink thickness is relatively thin. At this time, the brush pressure is 0.175MPa; brush wheel speed is 700r / min; board surface conveying speed is 2m / min; brushing time is 9s. When the gray value is 230-240+ standard deviation >20+ transparency ratio >10%, the ink thickness is judged to be 10-15µm, large area texture transparency, ink thickness is close to the critical value. At this time, the brush pressure is 0.13MPa; brush wheel speed is 500r / min; board surface conveying speed is 2.5m / min; brushing time is 5.5s. This achieves differentiated brushing treatment for reddish and non-reddish circuit boards. Before entering the pickling area, the circuit board is segmented into the smallest image acquisition units based on the image grid. When the number of reddish smallest units reaches 95% of the total number of smallest units, it is judged as "weak homogenization," and "short-time immersion pickling" is adopted in the pickling tank. If it is less than 95%, a "point-to-point spray pickling" scheme is adopted, and the pickling tank is set to a sulfuric acid concentration of 7%-9% and a temperature of 25-30℃. For reddish circuit boards judged as "weak homogenization," the immersion time in the pickling tank is ≤25s; for non-reddish circuit boards, the immersion time in the pickling tank is ≥4s. For the reddened circuit boards entering the spray pickling tank, based on the non-reddened area, reddened area, and different chromaticities of the reddened area determined by image segmentation, a differentiated spraying scheme is adopted: for non-reddened areas, a spraying pressure of 0.3MPa is used for 25s; for slightly reddened areas, a spraying pressure of 0.25MPa is used for 25s; for moderately reddened areas, a spraying pressure of 0.15MPa is used for 15s; and for heavily reddened areas, a spraying pressure of 0.12MPa is used for 15s, thereby achieving differentiated pickling operations. The circuit boards then enter the sandblasting chamber. For circuit boards without red tint, the same strategy as in the brush chamber is adopted, using a differentiated blasting scheme based on ink thickness. The sandblasting pressure and time decrease as the ink thickness decreases. For circuit boards with red tint, the impurity area is pre-treated with a reinforced sandblasting scheme before being uniformly subjected to rapid low-pressure sandblasting. This ensures that impurities are removed while improving the consistency of sanding the red tint circuit boards. After passing a unified quality inspection, the boards are sorted and unloaded.

[0031] In the entire solution, the vision-based approach is technology-driven, using image analysis methods to determine the ink characteristics of the circuit boards. Then, based on the preset plan, differentiated processing procedures are adopted for different characteristics of different circuit boards, so that all circuit boards can be processed in the same line. For the red-colored circuit boards, a synchronous processing procedure without grinding and rework is implemented, thereby improving the overall pass rate of circuit boards, reducing product damage, and eliminating the need to set up a complex processing line for the red-colored circuit boards, effectively saving costs.

[0032] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for non-grinding rework of a circuit board with a reddish-brown defect, characterized in that, Includes the following steps: S1. Loading: The vision module acquires images of the circuit board, and the images are transmitted to the analysis module for image analysis. S2. Extract relevant features based on image analysis, and generate control signals for brushing, pickling and sandblasting treatments based on the features respectively; S3. The circuit board enters the grinding, pickling and sandblasting processes in sequence, and the corresponding operation is performed according to the analysis results. The red-colored circuit board uses a non-grinding process; Before entering the pickling process, the area of ​​the reddened area on the circuit board is judged to determine whether to enter the spray pickling or immersion pickling tank for pickling treatment. When entering the sandblasting process, special targeted sandblasting treatment is adopted for the red-hot circuit boards; S4. Before unloading, the visual module is used to detect whether the material is qualified, and then the material is unloaded according to its classification.

2. The method for non-grinding rework of a circuit board with a reddish tint defect according to claim 1, characterized in that, In step S2, the features extracted after image analysis include the presence of redness, the ratio of the red area to the total area, the depth of redness, and abnormal color spot features. The operation is based on image features to control the brushing, pickling, and sandblasting processes. During image analysis, the presence of redness is determined by each smallest unit. When any smallest unit is red, it is determined to be a "red circuit board". At this time, a brushless solution is adopted in the brushing chamber. When the number of reddened smallest units reaches 95% of the total number of smallest units, it is judged as "weak homogenization" and "short-time immersion pickling" is adopted in the pickling tank. If it is less than 95%, "point-to-point spray pickling" is adopted. When concentrated spots are present in certain areas, and the number of the smallest units containing spots in the concentrated area reaches the minimum spot area threshold, the area is marked as an impurity area. During the sandblasting process, this area is subjected to additional concentrated sandblasting cleaning.

3. The method for non-grinding rework of a circuit board with a reddish tint defect according to claim 2, characterized in that, The brush chamber is equipped with liftable brushes, which move as the circuit board passes through the brush chamber. If it is determined to be a "red-light circuit board", the brush will rise to its highest point. At this time, the brush will not be in contact with the "red-light circuit board" and the brush will not need to be driven. If the circuit board is determined to be "non-red-emitting", the ink thickness is determined based on the strong correlation between the thickness of the solder resist ink and its surface gloss, grayscale value, and texture uniformity. The image analysis software removes dust and noise through image filtering, converts the color image to a grayscale image through grayscale processing, extracts the ink area and background through binarization segmentation, and then calculates the average grayscale value of all pixels in the detection area (value 0-255); extracts the standard deviation of gloss in the image; identifies the proportion of the transparent area of ​​the glass fiber texture on the board surface, and then controls the brushing component after judgment.

4. The method for non-grinding rework of a circuit board with a reddish tint defect according to claim 3, characterized in that, The execution logic for controlling the brush components includes: Grayscale value 180-220+ standard deviation <10+ transparency ratio <3%, ink thickness is judged to be 20-35μm, surface is flat, no texture is visible, ink thickness is standard thickness, at this time the brush pressure is 0.35MPa; brush wheel speed is 900r / min; board surface conveying speed is 1.5m / min; brushing time is 14s; Grayscale value 220-230+, standard deviation 10-20+, transparency ratio 3%-10%, ink thickness judged to be 15-20um, slight transparency in some areas, no reddish tinge, ink thickness is relatively thin, at this time the brush pressure is 0.175MPa; brush wheel speed 700r / min; board surface conveying speed 2m / min; brushing time 9s; Grayscale value 230-240+ standard deviation >20+ transparency ratio >10%, indicating ink thickness of 10-15um, large area texture transparency, ink thickness close to the critical value, at this time the brush pressure is 0.13MPa; brush wheel speed is 500r / min; board surface conveying speed is 2.5m / min; brushing time is 5.5s.

5. A method for non-grinding rework of a circuit board with a reddish tint defect according to claim 3, characterized in that, In step S3, a diversion scheme is adopted to transport the circuit board to the spray pickling tank and the immersion pickling tank through two conveyor belts respectively. A driving component is designed at the diversion point. The driving component control signal is generated based on the image feature analysis results so that the driving component can push the circuit board into the spray pickling tank or the immersion pickling tank. Since both red-tinted and non-red-tinted circuit boards will be immersed in the same pickling tank for pickling, the pickling tank is set with a sulfuric acid concentration of 7%-9% and a temperature of 25-30℃. For red-tinted circuit boards that are judged to have "weak homogenization", the immersion time in the pickling tank is ≤25s; for non-red-tinted circuit boards, the immersion time in the pickling tank is ≥45s.

6. The method for non-grinding rework of a circuit board with a reddish tint defect according to claim 5, characterized in that, For the reddened circuit boards entering the spray pickling tank, a differentiated spraying scheme is adopted based on the non-reddened area, reddened area, and different chromaticities of the reddened area determined by image segmentation, including: In non-reddening areas, a spray pressure of 0.3 MPa is used for a spraying time of 25 seconds; For areas with slight redness, spray at a pressure of 0.25 MPa for 25 seconds. For areas with moderate reddening, use a spray pressure of 0.15 MPa and a spray time of 15 seconds. For areas with severe redness, use a spray pressure of 0.12 MPa and a spray time of 15 seconds. By using multi-point nozzles for the spray area, all nozzles are supplied with high-pressure acid via a plunger pump, and pressure and spraying time are controlled by connecting precision pressure reducing valves and stop valves respectively.

7. A method for non-grinding rework of a circuit board with a reddish tint defect according to claim 3, characterized in that, When the sandblasting chamber performs sandblasting on circuit boards, for circuit boards without red glow, a differentiated spraying scheme is adopted based on the ink thickness, wherein the sandblasting pressure and sandblasting time decrease as the ink thickness decreases; For red-tinted circuit boards, after pre-treatment of impurity areas using a reinforced sandblasting method, a uniform rapid low-pressure sandblasting is performed to ensure impurity removal while improving the consistency of polishing the red-tinted circuit boards.

8. A method for non-grinding rework of a circuit board with a reddish tint defect according to claim 7, characterized in that, In step S4, another set of detection modules is set on the unloading conveyor belt at the outlet of the sandblasting chamber. Before unloading, the circuit board image is acquired by the vision detection module, and then the image analysis software performs image analysis to make a secondary judgment on the ink thickness of the circuit board. If the ink thickness is lower than the minimum threshold standard, it is judged as waste material; otherwise, it is judged as qualified material, and then it is sorted and unloaded.

9. A method for non-grinding rework of a circuit board with a reddish tint defect according to claim 2, characterized in that, In step S1, the vision module acquires an image of the circuit board surface below it, and the image is quickly analyzed by image analysis software to extract image features.

10. A method for non-grinding rework of a circuit board with a reddish tint defect according to claim 9, characterized in that, In the image processing process, the image segmentation method is used to divide the image into multiple rectangular regions. Then, the grid method is used to divide the smallest image acquisition unit. In the colorimetric judgment process, multiple intervals are divided by setting the colorimetric threshold, so as to judge the degree of oxidation of the resin glass fiber based on the colorimetric value. By extracting abnormal pattern spots, it can be determined that there is an accumulation of impurities.