Screen foreign matter dust inspection system and method based on multi-angle camera scanning

Abstract

The invention provides a screen foreign matter and dust inspection system and method based on multi-angle camera scanning, and relates to the technical field of screen detection.The method comprises the steps that a multi-angle scanning module conducts image scanning on a to-be-detected screen to obtain a plurality of sample images; segmenting each sample image into a plurality of image blocks and extracting a block gray value in each image block; and for each image block, respectively calculating a difference value between the block gray value of the image block and the block gray value of each adjacent image block, and when the difference value is greater than a preset parameter, carrying out line-by-line pixel analysis on the image block as a foreign matter image block to extract foreign matter information in the foreign matter image block and recording the foreign matter information. The beneficial effects are that the multi-angle scanning module can obtain screen images from multiple angles, and detection blind areas possibly caused by a single angle are avoided; compared with manual visual inspection, the system can keep a highly consistent detection standard, so that the stability and reliability of an inspection result are ensured.

Description

Technical Field

[0001] The present invention relates to the field of screen detection technology, and in particular to a screen foreign matter dust inspection system and method based on multi-angle camera scanning. Background Art

[0002] At present, the inspection of large-size LCD screens is mainly carried out by manual visual inspection. This traditional method has exposed a series of significant shortcomings in practical applications. First of all, from the perspective of time cost, the inspection process of large-size LCD screens takes a long time. Due to the increase in screen size, inspectors need to carefully examine every corner of the screen to ensure that there are no flaws or defects. This meticulous and cumbersome process undoubtedly greatly prolongs the production cycle and reduces the overall production efficiency.

[0003] Secondly, high labor costs are also a major pain point of existing technologies. Due to the complexity of the inspection work and the requirements for precision, several inspectors are often required to intervene at the same time to ensure the comprehensiveness and accuracy of the inspection. This not only increases the labor costs of enterprises, but also limits the expansion of production scale to a certain extent. In addition, with the changes in the labor market, the continuous increase in labor costs has further exacerbated this problem.

[0004] More importantly, manual visual inspection is carried out with the LCD screen lit, and inspectors stare at the screen for a long time, which is very easy to cause visual fatigue. This fatigue state not only affects the judgement and concentration of inspectors, but also may lead to the omission of defective products, which seriously affects the quality of products and customer satisfaction. In modern manufacturing, product quality is the cornerstone of enterprise survival and development, and any negligence may bring immeasurable losses to enterprises.

[0005] In summary, the existing production lines have shortcomings in large-size LCD screen inspection, such as long time, high labor costs, and easy errors. These problems not only restrict the improvement of production efficiency, but also pose a serious threat to the company's cost control and product quality. Therefore, seeking a more efficient, accurate, and low-cost inspection method has become an important issue that needs to be solved in the current large-size LCD screen production process. Summary of the invention

[0006] In view of the problems existing in the prior art, the present invention provides a screen foreign matter dust inspection system based on multi-angle camera scanning, comprising:

[0007] A multi-angle scanning module is directed toward the screen to be inspected, and is used to scan the screen to be inspected to obtain multiple sample images;

[0008] An image analysis module, connected to the multi-angle scanning module, for dividing each of the sample images into a plurality of image blocks and extracting a block gray value in each of the image blocks;

[0009] A foreign body extraction module is connected to the image analysis module and is used to calculate the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block. When the difference is greater than a preset parameter, the image block is treated as a foreign body image block and pixel analysis is performed row by row to extract the foreign body information therein and record it.

[0010] Preferably, the image analysis module includes:

[0011] A boundary removal unit, used for performing boundary removal processing on each of the sample images;

[0012] An image segmentation unit, connected to the boundary removal unit, and configured to segment each of the sample images after boundary processing into a plurality of image blocks according to a set image block size;

[0013] The gray value extraction unit is connected to the image segmentation unit and is used for taking the gray value accumulation value of all pixels in each image block as the corresponding block gray value.

[0014] Preferably, the foreign matter extraction module comprises:

[0015] A transparency extraction unit, used for taking the average gray value of each pixel in the foreign body image block as transparency;

[0016] A length and width extraction unit is used to traverse each pixel point in the foreign body image block, and take each pixel point that is different from the marked pixel point as a foreign body pixel point, and then take the distance between the two foreign body pixel points that are farthest in the length direction as the foreign body length, and take the distance between the two foreign body pixel points that are farthest in the width direction as the foreign body width;

[0017] An area extraction unit, connected to the length and width extraction unit, and used to take the total area of ​​each of the foreign object pixel points as the foreign object area;

[0018] A morphology extraction unit, connected to the length and width extraction unit, is used to extract the discreteness and shape of the image surrounded by each of the foreign object pixel points as the foreign object morphology, and calculate the percentage of foreign object pixels of each of the foreign object pixel points in the foreign object image block as the distribution uniformity;

[0019] The information recording unit is respectively connected to the transparency extraction unit, the length and width extraction unit, the area extraction unit and the morphology extraction unit, and is used to save the transparency, the foreign body length, the foreign body width, the foreign body area, the foreign body morphology and the distribution uniformity ratio as the foreign body information.

[0020] Preferably, the multi-angle scanning module includes a plurality of camera groups arranged in rows, and each of the camera groups includes a plurality of cameras facing in different angles.

[0021] Preferably, each camera group includes two oblique view cameras and one straight view camera, the oblique view cameras are used to obtain oblique view images, and the straight view cameras are used to obtain straight view images, and the foreign body extraction module further includes:

[0022] A light flux judgment unit is used to calculate the strabismus light flux of the strabismus image and the straight light flux of the straight-view image for the same group of the camera groups, and to record the strabismus light flux as poor when the straight-view light flux is good and the strabismus light flux is poor, to record the positive light flux as poor when the straight-view light flux is poor and the strabismus light flux is good, and to record the full light flux as poor when the straight-view light flux is poor and the strabismus light flux is poor.

[0023] The present invention also provides a screen foreign matter gray inspection method based on multi-angle camera scanning, which is applied to the above-mentioned screen foreign matter gray inspection system, comprising:

[0024] Step S1, the screen foreign body grayscale inspection system performs image scanning on the screen to be inspected to obtain multiple sample images;

[0025] Step S2, the screen foreign matter grayscale inspection system divides each of the sample images into a plurality of image blocks and extracts a block grayscale value in each of the image blocks;

[0026] Step S3, the screen foreign body grayscale inspection system calculates the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block, and determines whether the difference is greater than a preset parameter:

[0027] If yes, the image block is treated as a foreign body image block, and pixel analysis is performed row by row to extract and record the foreign body information therein;

[0028] If not, the process returns to step S1 to perform image scanning on the next screen to be detected.

[0029] Preferably, step S2 comprises:

[0030] Step S21, the screen foreign matter grayscale inspection system performs a de-bordering process on each of the sample images;

[0031] Step S22, the screen foreign matter grayscale inspection system divides each of the sample images after boundary processing into a plurality of image blocks according to a set image block size;

[0032] Step S23, the screen foreign matter grayscale inspection system uses, for each image block, the accumulated value of the grayscale values ​​of all pixels therein as the corresponding block grayscale value.

[0033] Preferably, the process of extracting the foreign body information in step S3 includes:

[0034] Step S31, the screen foreign body grayscale inspection system uses the average grayscale value of each pixel in the foreign body image block as transparency;

[0035] Step S32, the screen foreign body grayscale inspection system traverses each pixel point in the foreign body image block, and takes each pixel point that is different from the marked pixel point as a foreign body pixel point, and then takes the distance between the two foreign body pixels farthest in the length direction as the foreign body length, and takes the distance between the two foreign body pixels farthest in the width direction as the foreign body width;

[0036] Step S33, the screen foreign body grayscale inspection system uses the total area of ​​each foreign body pixel as the foreign body area;

[0037] Step S34, the screen foreign body grayscale inspection system extracts the discreteness and shape of the image surrounded by each of the foreign body pixel points as the foreign body morphology, and calculates the percentage of foreign body pixels of each of the foreign body pixel points in the foreign body image block as the distribution uniformity;

[0038] In step S35, the screen foreign matter grayscale inspection system saves the transparency, the foreign matter length, the foreign matter width, the foreign matter area, the foreign matter shape and the distribution uniformity ratio as the foreign matter information.

[0039] Preferably, the multi-angle scanning module includes a plurality of camera groups arranged in rows, and each of the camera groups includes a plurality of cameras facing in different angles.

[0040] Preferably, each camera group includes two oblique-view cameras and one straight-view camera, the oblique-view cameras are used to capture oblique images, and the straight-view cameras are used to capture straight-view images. Then, step S3 further includes a light flux determination process:

[0041] The screen foreign matter grayscale inspection system calculates the strabismus light transmittance of the strabismus image and the straight light transmittance of the straight-view image for the same group of the camera groups, respectively, and records the strabismus light transmittance as poor when the straight light transmittance is good and the strabismus light transmittance is poor, records the positive light transmittance as poor when the straight light transmittance is poor and the strabismus light transmittance is good, and records the full light transmittance as poor when the straight light transmittance is poor and the strabismus light transmittance is poor.

[0042] The above technical solution has the following advantages or beneficial effects:

[0043] 1. The multi-angle scanning module can obtain screen images from multiple angles, avoiding the detection blind spots that may be caused by a single angle. At the same time, the image analysis module and foreign body extraction module can accurately extract foreign body information in the image through sophisticated image processing algorithms, greatly reducing the probability of missed detection and false detection.

[0044] 2. Compared with manual visual inspection, the system can maintain highly consistent inspection standards. Regardless of the inspector's experience level, the system can make accurate judgments according to preset parameters and algorithms, thereby ensuring the stability and reliability of the inspection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The structure diagram of a screen foreign matter dust inspection system based on multi-angle camera scanning in a preferred embodiment of the present invention;

[0046] Figure 2 This is a structural schematic diagram of a multi-angle scanning module in a preferred embodiment of the present invention;

[0047] Figure 3 It is a structural schematic diagram of a camera group in a preferred embodiment of the present invention;

[0048] Figure 4 The figure is a flow chart of a method for inspecting screen foreign matter based on multi-angle camera scanning in a preferred embodiment of the present invention;

[0049] Figure 5 Schematic diagram of a sub-process of step S2 in a preferred embodiment of the present invention;

[0050] Figure 6 This is a schematic diagram of a sub-process of step S3 in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0051] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. The present invention is not limited to this embodiment, and other embodiments may also fall within the scope of the present invention as long as they conform to the gist of the present invention.

[0052] In a preferred embodiment of the present invention, based on the above problems existing in the prior art, a screen foreign matter dust inspection system based on multi-angle camera scanning is provided. Figure 1 As shown, including:

[0053] A multi-angle scanning module 1 is directed toward the screen to be detected and is used to scan the screen to be detected to obtain multiple sample images;

[0054] The image analysis module 2 is connected to the multi-angle scanning module 1 and is used to divide each sample image into multiple image blocks and extract the block gray value in each image block;

[0055] The foreign body extraction module 3 is connected to the image analysis module 2 and is used to calculate the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block. When the difference is greater than a preset parameter, the image block is treated as a foreign body image block and pixel analysis is performed row by row to extract the foreign body information therein and record it.

[0056] Specifically, the present embodiment provides a screen foreign matter and dust inspection system based on multi-angle camera scanning, which provides an effective solution to the deficiencies of the prior art in the inspection of large-size LCD screens.

[0057] Through the multi-angle scanning module 1, the system can quickly scan the large screen in all directions without blind spots and obtain multiple sample images. This parallel processing method greatly improves the scanning efficiency and greatly shortens the inspection time compared to manual visual inspection.

[0058] The image analysis module 2 and the foreign matter extraction module 3 can automatically process and analyze the scanned images without manual intervention, further shortening the overall inspection cycle.

[0059] The entire inspection process is highly automated, eliminating the need for multiple inspectors to intervene simultaneously. The system can independently complete tasks such as image scanning, analysis, and foreign body extraction, greatly reducing reliance on manual labor.

[0060] The multi-angle scanning module 1 can obtain screen images from multiple angles, avoiding the detection blind spots that may be caused by a single angle. At the same time, the image analysis module and the foreign body extraction module can accurately extract the foreign body information in the image through sophisticated image processing algorithms, greatly reducing the probability of missed detection and false detection.

[0061] Compared with manual visual inspection, the system can maintain highly consistent inspection standards. Regardless of the inspector's experience level, the system can make accurate judgments according to preset parameters and algorithms, thus ensuring the stability and reliability of the inspection results.

[0062] In summary, this screen foreign matter dust inspection system based on multi-angle camera scanning effectively solves the shortcomings of the existing technology in large-size LCD screen inspection, such as long time, high labor cost, and easy error, through efficient scanning, automated analysis, and high-precision detection. This technical solution not only improves production efficiency, but also reduces enterprise costs, while ensuring product quality and customer satisfaction.

[0063] In a preferred embodiment of the present invention, the image analysis module 2 includes:

[0064] A de-bordering unit 21, used for performing de-bordering processing on each sample image;

[0065] The image segmentation unit 22 is connected to the boundary removal unit 21 and is used to segment each sample image after boundary processing into a plurality of image blocks according to a set image block size;

[0066] The gray value extraction unit 23 is connected to the image segmentation unit 22 and is used for taking the gray value accumulation value of all the pixels in each image block as the corresponding block gray value.

[0067] Specifically, in this embodiment, in actual screen detection, the edge of the screen may produce some image features that are different from the inside of the screen due to light, reflection, or errors in the manufacturing process. If these features are not processed, they may interfere with subsequent image analysis and foreign body detection. Specifically, in this embodiment, the edge of the luminous LCD screen is obtained by the scanning start point of the multi-angle scanning module and the corresponding boundary position is obtained. The boundary removal unit can eliminate these interference factors by removing the edge of the image to ensure the accuracy of subsequent analysis.

[0068] Splitting the image after boundary processing into multiple image blocks according to the set image block size is a common strategy in image processing. The advantage of this is that complex image problems can be converted into smaller and easier to handle problems. At the same time, by splitting the image blocks, each image block can be processed in parallel to improve the overall processing efficiency. In addition, the size of the image block can be adjusted according to actual needs to meet the needs of screen detection of different sizes and resolutions.

[0069] Grayscale value is an indicator to measure the brightness of an image. For screen foreign body detection, the difference in grayscale value is often an important basis for judging the presence of foreign bodies. The grayscale value extraction unit can simplify the subsequent foreign body detection algorithm and improve the detection accuracy by calculating the accumulated grayscale value of all pixels in each image block as the block grayscale value. At the same time, since the accumulated grayscale value is a scalar value, it can greatly reduce the amount of calculation and storage requirements compared to directly processing the grayscale value of each pixel.

[0070] In summary, the image analysis module in this embodiment solves the problem of image boundary interference, improves the efficiency and accuracy of image processing, and provides more accurate and reliable image data for subsequent foreign body detection. These improvements make the entire screen foreign body detection system more stable, efficient and accurate in practical applications.

[0071] In a preferred embodiment of the present invention, the foreign matter extraction module 3 includes:

[0072] A transparency extraction unit 31 is used to use the average gray value of each pixel in the foreign body image block as transparency;

[0073] The length and width extraction unit 32 is used to traverse each pixel point in the foreign body image block, and take each pixel point different from the marked pixel point as a foreign body pixel point, and then take the distance between the two foreign body pixels farthest in the length direction as the foreign body length, and take the distance between the two foreign body pixels farthest in the width direction as the foreign body width;

[0074] An area extraction unit 33, connected to the length and width extraction unit 32, is used to take the total area of ​​each foreign object pixel as the foreign object area;

[0075] The morphology extraction unit 34 is connected to the length and width extraction unit 32, and is used to extract the discreteness and shape of the image surrounded by each foreign object pixel as the foreign object morphology, and calculate the foreign object pixel percentage of each foreign object pixel in the foreign object image block as the distribution uniformity;

[0076] The information recording unit 35 is respectively connected to the transparency extraction unit 31, the length and width extraction unit 32, the area extraction unit 33 and the morphology extraction unit 34, and is used to save the transparency, foreign body length, foreign body width, foreign body area, foreign body morphology and distribution uniformity ratio as foreign body information.

[0077] Specifically, in screen detection, the transparency of foreign objects is often an important feature. The transparency extraction unit can quantify the transparency characteristics of foreign objects by calculating the average gray value of each pixel in the foreign object image block as the transparency, providing a basis for subsequent analysis and judgment.

[0078] The length and width extraction unit and the area extraction unit calculate the length, width and area of ​​the foreign object respectively. These size characteristics are important bases for judging the type and severity of the foreign object. For example, a tiny foreign object may not affect the use of the screen, while a larger foreign object may need to be repaired or replaced.

[0079] The morphological extraction unit not only extracts the discreteness and shape characteristics of foreign objects, but also calculates the percentage of foreign object pixels in the foreign object image block as the distribution uniformity. The discreteness of foreign objects is used to determine whether the foreign objects are discrete or concentrated. If they are concentrated, it can be further determined whether their shapes are circular, strip-shaped, or polygonal. These features can provide more detailed foreign object information and help determine the nature, source, and possible causes of foreign objects.

[0080] By extracting multiple features of foreign matter (such as transparency, size, shape and distribution uniformity) and saving these features as foreign matter information, the accuracy and reliability of foreign matter detection can be greatly improved. These feature information can provide strong support for subsequent screen repair, quality control and fault analysis.

[0081] Furthermore, multiple feature information of foreign matter is saved in a structured manner to facilitate subsequent processing and analysis. For example, the screen can be classified, counted and predicted based on the feature information of foreign matter, providing data support for the optimization and improvement of the production line.

[0082] In a preferred embodiment of the present invention, the multi-angle scanning module 1 includes a plurality of camera groups arranged in rows, each camera group includes a plurality of cameras facing at different angles. Specifically, each camera group includes two oblique cameras 11 and one straight camera 12, the oblique camera 11 captures an oblique image, and the straight camera 12 captures a straight image, and the foreign body extraction module 3 further includes:

[0083] The light transmittance judgment unit 36 ​​is used to calculate the strabismus light transmittance of the strabismus image and the straight light transmittance of the straight-view image for the same group of cameras, and to record the strabismus light transmittance as poor when the straight-view light transmittance is good and the strabismus light transmittance is poor, to record the positive light transmittance as poor when the straight-view light transmittance is poor and the strabismus light transmittance is good, and to record the full light transmittance as poor when the straight-view light transmittance is poor and the strabismus light transmittance is poor.

[0084] Specifically, Figure 2 and Figure 3 As shown, the camera groups arranged in rows each include three cameras, and their orientations are respectively perpendicular to the screen to be detected, inclined 45° to the left side of the screen to be detected, inclined 45° to the right side of the screen to be detected, or perpendicular to the screen to be detected, inclined 45° to the width direction of the screen to be detected, and inclined 45° to the length direction of the screen to be detected. Figure 3 The figure shows the tilting method of the two. By arranging multiple groups along the width direction of the screen to be detected, the entire screen to be detected can be covered. Furthermore, a whole row of camera groups can be moved from one side of the screen to be detected to the other side under the drive of the cylinder, and the screen to be detected can be photographed line by line. Although it will increase the detection time to a certain extent, it can perform comprehensive screen detection.

[0085] By arranging several camera groups in rows, each group contains multiple cameras facing at different angles (such as two oblique cameras and one straight camera), multi-angle scanning of the screen can be achieved. This multi-angle scanning can capture details and anomalies in different areas of the screen, improving the comprehensiveness of detection.

[0086] Luminous flux judgment: Luminous flux refers to the ratio or efficiency of light passing through the screen, and is an important indicator for evaluating screen quality. The luminous flux judgment unit calculates the luminous flux of the oblique image and the direct luminous flux of the direct image respectively, which can more accurately judge the luminous flux of the screen.

[0087] Poor squinting light transmittance: When the direct view light transmittance is good, but the squinting light transmittance is poor, it may indicate that there are defects such as tiny scratches, dents or bumps on the screen surface, which are not easy to detect when looking straight, but become obvious when looking at the side.

[0088] Poor positive luminous flux: When the direct luminous flux is poor, but the squinting luminous flux is good, it may indicate that there are defects inside the screen, such as internal cracks, bubbles or impurities. These defects are not easy to capture when squinting, but will significantly affect the screen display effect when looking straight.

[0089] Poor total luminous flux: When both the direct view luminous flux and the oblique view luminous flux are poor, it indicates that the screen has serious luminous flux problems and may need to be replaced or repaired.

[0090] Through multi-angle scanning and luminous flux judgment, the luminous flux problem of the screen can be quickly located, reducing the time and cost of manual inspection. Based on the comparison results of the luminous flux of the strabismus and the luminous flux of the straight view, the severity of the luminous flux problem of the screen can be accurately evaluated, providing a decision-making basis for subsequent repair or replacement.

[0091] By testing the light transmittance of the screen, problems in the production process can be discovered in a timely manner, the production process can be optimized, and product quality and consistency can be improved.

[0092] The present invention also provides a screen foreign body gray inspection method based on multi-angle camera scanning, which is applied to the above-mentioned screen foreign body gray inspection system, such as Figure 4 As shown, including:

[0093] Step S1, the screen foreign body grayscale inspection system performs image scanning on the screen to be inspected to obtain multiple sample images;

[0094] Step S2, the screen foreign body grayscale inspection system divides each sample image into multiple image blocks and extracts the block grayscale value in each image block;

[0095] Step S3, the screen foreign body grayscale inspection system calculates the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block, and determines whether the difference is greater than a preset parameter:

[0096] If yes, the image block is treated as a foreign body image block, and pixel analysis is performed row by row to extract the foreign body information therein and record it;

[0097] If not, the process returns to step S1 to perform image scanning on the next screen to be inspected.

[0098] In a preferred embodiment of the present invention, Figure 5 As shown, step S2 includes:

[0099] Step S21, the screen foreign matter grayscale inspection system performs boundary removal processing on each sample image;

[0100] Step S22, the screen foreign matter grayscale inspection system divides each sample image after boundary processing into multiple image blocks according to a set image block size;

[0101] Step S23, the screen foreign matter grayscale inspection system uses the accumulated grayscale values ​​of all pixels in each image block as the corresponding block grayscale value.

[0102] In a preferred embodiment of the present invention, Figure 6 As shown, the process of extracting foreign matter information in step S3 includes:

[0103] Step S31, the screen foreign body grayscale inspection system uses the average grayscale value of each pixel in the foreign body image block as transparency;

[0104] Step S32, the screen foreign body grayscale inspection system traverses each pixel point in the foreign body image block, and takes each pixel point that is different from the marked pixel point as a foreign body pixel point, and then takes the distance between the two farthest foreign body pixel points in the length direction as the foreign body length, and takes the distance between the two farthest foreign body pixel points in the width direction as the foreign body width;

[0105] Step S33, the screen foreign body grayscale inspection system uses the total area of ​​each foreign body pixel as the foreign body area;

[0106] Step S34, the screen foreign body grayscale inspection system extracts the discreteness and shape of the image surrounded by each foreign body pixel as the foreign body morphology, and calculates the percentage of foreign body pixels of each foreign body pixel in the foreign body image block as the distribution uniformity;

[0107] In step S35, the screen foreign matter grayscale inspection system saves the transparency, foreign matter length, foreign matter width, foreign matter area, foreign matter shape and distribution uniformity ratio as foreign matter information.

[0108] In a preferred embodiment of the present invention, the multi-angle scanning module includes a plurality of camera groups arranged in rows, and each camera group includes a plurality of cameras facing in different angles.

[0109] In a preferred embodiment of the present invention, each camera group includes two oblique cameras and one straight camera, the oblique cameras are used to capture oblique images, and the straight cameras are used to capture straight images. Then step S3 also includes a light flux determination process:

[0110] For the same group of cameras, the screen foreign matter grayscale inspection system calculates the strabismus light transmittance of the strabismus image and the straight light transmittance of the straight-view image respectively, and records the strabismus light transmittance as poor when the straight-view light transmittance is good and the strabismus light transmittance is poor, records the positive light transmittance as poor when the straight-view light transmittance is poor and the strabismus light transmittance is good, and records the full light transmittance as poor when the straight-view light transmittance is poor and the strabismus light transmittance is poor.

[0111] The above are only preferred embodiments of the present invention, and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the contents of this specification and illustrations should be included in the protection scope of the present invention.

Claims

1. A screen foreign matter dust inspection system based on multi-angle camera scanning, characterized in that: include: A multi-angle scanning module is directed toward the screen to be inspected, and is used to scan the screen to be inspected to obtain multiple sample images; An image analysis module, connected to the multi-angle scanning module, for dividing each of the sample images into a plurality of image blocks and extracting a block gray value in each of the image blocks; A foreign body extraction module is connected to the image analysis module and is used to calculate the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block. When the difference is greater than a preset parameter, the image block is treated as a foreign body image block and pixel analysis is performed row by row to extract the foreign body information therein and record it.

2. The screen foreign matter dust inspection system according to claim 1, characterized in that: Image analysis modules include: A boundary removal unit, used for performing boundary removal processing on each of the sample images; An image segmentation unit, connected to the boundary removal unit, and configured to segment each of the sample images after boundary processing into a plurality of image blocks according to a set image block size; The gray value extraction unit is connected to the image segmentation unit and is used for taking the gray value accumulation value of all pixels in each image block as the corresponding block gray value.

3. The screen foreign matter dust inspection system according to claim 1, characterized in that: The foreign body extraction module comprises: A transparency extraction unit, used for taking the average gray value of each pixel in the foreign body image block as transparency; A length and width extraction unit is used to traverse each pixel point in the foreign body image block, and take each pixel point that is different from the marked pixel point as a foreign body pixel point, and then take the distance between the two foreign body pixel points that are farthest in the length direction as the foreign body length, and take the distance between the two foreign body pixel points that are farthest in the width direction as the foreign body width; An area extraction unit, connected to the length and width extraction unit, and used to take the total area of ​​each of the foreign object pixel points as the foreign object area; A morphology extraction unit, connected to the length and width extraction unit, is used to extract the discreteness and shape of the image surrounded by each of the foreign object pixel points as the foreign object morphology, and calculate the percentage of foreign object pixels of each of the foreign object pixel points in the foreign object image block as the distribution uniformity; The information recording unit is respectively connected to the transparency extraction unit, the length and width extraction unit, the area extraction unit and the morphology extraction unit, and is used to save the transparency, the foreign body length, the foreign body width, the foreign body area, the foreign body morphology and the distribution uniformity ratio as the foreign body information.

4. The screen foreign matter dust inspection system according to claim 1, characterized in that: The multi-angle scanning module includes a plurality of camera groups arranged in rows, and each camera group includes a plurality of cameras facing in different angles.

5. The screen foreign matter dust inspection system according to claim 4, characterized in that: Each camera group includes two oblique view cameras and one straight view camera. The oblique view cameras take pictures to obtain oblique view images, and the straight view cameras take pictures to obtain straight view images. Then, the foreign body extraction module further includes: A light flux judgment unit is used to calculate the strabismus light flux of the strabismus image and the straight light flux of the straight-view image for the same group of the camera groups, and to record the strabismus light flux as poor when the straight-view light flux is good and the strabismus light flux is poor, to record the positive light flux as poor when the straight-view light flux is poor and the strabismus light flux is good, and to record the full light flux as poor when the straight-view light flux is poor and the strabismus light flux is poor.

6. A screen foreign matter dust inspection method based on multi-angle camera scanning, characterized in that: A screen foreign body grayscale inspection system as claimed in any one of claims 1 to 5, comprising: Step S1, the screen foreign body grayscale inspection system performs image scanning on the screen to be inspected to obtain multiple sample images; Step S2, the screen foreign matter grayscale inspection system divides each of the sample images into a plurality of image blocks and extracts a block grayscale value in each of the image blocks; Step S3, the screen foreign body grayscale inspection system calculates the difference between the block grayscale value of each image block and the block grayscale values ​​of each adjacent image block, and determines whether the difference is greater than a preset parameter: If yes, the image block is treated as a foreign body image block, and pixel analysis is performed row by row to extract and record the foreign body information therein; If not, the process returns to step S1 to perform image scanning on the next screen to be detected.

7. The screen foreign matter dust inspection method according to claim 6, characterized in that: The step S2 comprises: Step S21, the screen foreign matter grayscale inspection system performs a de-bordering process on each of the sample images; Step S22, the screen foreign matter grayscale inspection system divides each of the sample images after boundary processing into a plurality of image blocks according to a set image block size; Step S23, the screen foreign matter grayscale inspection system uses, for each image block, the accumulated value of the grayscale values ​​of all pixels therein as the corresponding block grayscale value.

8. The screen foreign matter dust inspection method according to claim 6, characterized in that: The process of extracting the foreign body information in step S3 includes: Step S31, the screen foreign body grayscale inspection system uses the average grayscale value of each pixel in the foreign body image block as transparency; Step S32, the screen foreign body grayscale inspection system traverses each pixel point in the foreign body image block, and takes each pixel point that is different from the marked pixel point as a foreign body pixel point, and then takes the distance between the two foreign body pixels farthest in the length direction as the foreign body length, and takes the distance between the two foreign body pixels farthest in the width direction as the foreign body width; Step S33, the screen foreign body grayscale inspection system uses the total area of ​​each foreign body pixel as the foreign body area; Step S34, the screen foreign body grayscale inspection system extracts the discreteness and shape of the image surrounded by each of the foreign body pixel points as the foreign body morphology, and calculates the percentage of foreign body pixels of each of the foreign body pixel points in the foreign body image block as the distribution uniformity; In step S35, the screen foreign matter grayscale inspection system saves the transparency, the foreign matter length, the foreign matter width, the foreign matter area, the foreign matter shape and the distribution uniformity ratio as the foreign matter information.

9. The screen foreign matter dust inspection method according to claim 6, characterized in that: The multi-angle scanning module includes a plurality of camera groups arranged in rows, and each camera group includes a plurality of cameras facing in different angles.

10. The screen foreign matter dust inspection method according to claim 9, characterized in that: Each camera group includes two oblique cameras and one straight-view camera, the oblique cameras capture oblique images, and the straight-view cameras capture straight-view images, then the step S3 also includes a luminous flux judgment process: the screen foreign matter grayscale inspection system calculates the oblique luminous flux of the oblique image and the straight-view image for the same camera group, respectively, and records the oblique luminous flux as poor when the straight-view luminous flux is good and the oblique luminous flux is poor, records the positive luminous flux as poor when the straight-view luminous flux is poor and the oblique luminous flux is good, and records the full luminous flux as poor when the straight-view luminous flux is poor and the oblique luminous flux is poor.

Images