A mold part quality control system

Abstract

The application discloses a kind of mould parts quality control systems, including camera module, image processing module and quality control module;Camera module is used to obtain the detection image of mould parts;Image processing module is used to judge whether mould parts pass quality inspection based on detection image, including the gray scale processing of detection image, obtains gray scale image;The foreground detection of gray scale image is obtained, and foreground image is obtained;After noise reduction processing to foreground image, judge whether mould parts pass quality inspection;Quality control module is used to the quantity of mould parts that do not pass quality inspection is counted, obtains statistical result, and is used to the quality control of mould parts based on statistical result.The application effectively reduces the number of pixel points that enters into noise reduction processing stage, to improve the efficiency of quality inspection to mould parts using machine vision, it is beneficial to quickly quality detection to mould parts.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of quality control, and particularly relates to a mold part quality control system. BACKGROUND

[0002] Mold, that is, various molds and tools used in industrial production to obtain required products through injection molding, blow molding, extrusion, die casting or forging, smelting, stamping and the like. The mold is a tool for making shaped objects, and the tool is composed of various mold parts, and different molds are composed of different parts. The mold mainly realizes the processing of the shape of the product by changing the physical state of the molding material.

[0003] After the mold parts are produced, quality inspection needs to be performed. With the development of technology, the way of quality inspection of mold parts has changed from traditional manual quality inspection to machine vision quality inspection.

[0004] In the existing visual inspection process, in the process of denoising the obtained mold part image, all pixel points in the image are directly processed, and many useless background pixel points are also denoised. Such processing method affects the efficiency of visual inspection and is not conducive to rapid quality inspection of mold parts. SUMMARY

[0005] The purpose of the present application is to disclose a mold part quality control system, which solves the problem that in the prior art, when machine vision is used to detect the quality of mold parts, all pixel points in the image are directly denoised, which affects the efficiency of visual inspection and is not conducive to rapid quality inspection of mold parts.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0007] A mold part quality control system, comprising a camera module, an image processing module and a quality control module;

[0008] The camera module is used to obtain a detection image of the mold part;

[0009] The image processing module is used to determine whether the mold part passes the quality inspection based on the detection image;

[0010] The quality control module is used to count the number of mold parts that do not pass the quality inspection, obtain a statistical result, and perform quality control on the mold parts based on the statistical result;

[0011] The determination of whether the mold part passes the quality inspection based on the detection image comprises:

[0012] The detection image is subjected to a grayscale processing to obtain a grayscale image;

[0013] The grayscale image is subjected to a foreground detection to obtain a foreground image;

[0014] The foreground image is subjected to a noise reduction processing to determine whether the mold part passes the quality inspection.

[0015] Preferably, the camera module comprises a camera and a fill light;

[0016] The camera is configured to acquire the detection image of the mold part;

[0017] The fill light is configured to provide the camera with illuminating light.

[0018] Preferably, the image processing module comprises a storage unit, a grayscale unit, a foreground detection unit and an image recognition unit;

[0019] The storage unit is configured to store a comparison image;

[0020] The grayscale unit is configured to respectively subject the detection image and the comparison image to a grayscale processing to obtain a grayscale image and a grayscale comparison image;

[0021] The foreground detection unit is configured to subject the grayscale image to a foreground detection based on the grayscale comparison image to obtain a foreground image;

[0022] The image recognition unit is configured to subject the foreground image to a noise reduction processing to determine whether the mold part passes the quality inspection.

[0023] Preferably, the comparison image is different from the detection image in shooting time, and the comparison image is an image shot when the mold part is outside the shooting range of the camera module.

[0024] Preferably, the grayscale processing comprises:

[0025] For the image P, the grayscale processing is performed on P by using the following formula:

[0026] grayP(x,y)=w1×R(x,y)+w2×G(x,y)+w3×B(x,y)

[0027] In the formula, grayP represents the image obtained by subjecting the image P to the grayscale processing; grayP(x,y) represents the grayscale value of the pixel point with the coordinate (x,y) in grayP; R, G and B respectively represent the images corresponding to the red component, the green component and the blue component of the image P in the RGB color space; R(x,y), G(x,y) and B(x,y) respectively represent the pixel values of the pixel points with the coordinates (x,y) in R, G and B; w1, w2 and w3 represent preset weighting coefficients.

[0028] Preferably, the grayscale contrast image is used to perform foreground detection on the grayscale image to obtain a foreground image, including:

[0029] For a pixel point Pixel in the grayscale image, the judgment value of Pixel is calculated by the following formula:

[0030] judval(Pixel) = |grayimg(Pixel) - graycontimg(Pixel)|

[0031] In the formula, judval(Pixel) represents the judgment value of the pixel point Pixel, grayimg(Pixel) represents the grayscale value of the pixel point Pixel in the grayscale image grayimg, and graycontimg(Pixel) represents the grayscale value of the corresponding pixel point of Pixel in the grayscale contrast image.

[0032] If the judgment value is greater than a set judgment value threshold, the pixel point Pixel in the grayscale image is stored in a foreground set frtviwset.

[0033] The foreground image is obtained based on the pixel points in the foreground set frtviwset.

[0034] Preferably, the quality control module includes a statistical unit, a calculation unit, and a warning unit.

[0035] The statistical unit is configured to count the number of mold parts of the same production batch that fail the quality inspection to obtain a statistical result.

[0036] The calculation unit is configured to calculate a pass rate of the mold parts of the same production batch based on the statistical result.

[0037] The warning unit is configured to issue a warning prompt to a manager when the pass rate is less than a set pass rate threshold.

[0038] Preferably, the pass rate of the mold parts of the same production batch is calculated based on the statistical result, including:

[0039] The pass rate of the mold parts is calculated as follows:

[0040]

[0041] In the formula, pasrat represents the pass rate, num1 represents the number of mold parts of the same production batch that fail the quality inspection, and num2 represents the total number of mold parts of the same production batch.

[0042] The application first obtains the foreground image corresponding to the detection image of the mold part when using machine vision to perform quality inspection on the mold part, then performs noise reduction processing on the foreground image, and then judges whether the mold part passes the quality inspection based on the result of the noise reduction processing, effectively reducing the number of pixel points entering the noise reduction processing stage, thereby improving the efficiency of using machine vision to perform quality inspection on the mold part, and facilitating rapid quality detection of the mold part. BRIEF DESCRIPTION OF DRAWINGS

[0043] The application is further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the application. For ordinary skilled in the art, other drawings can be obtained without creative labor according to the following drawings.

[0044] Figure 1 An exemplary embodiment of a mold part quality control system of the application.

[0045] Figure 2 An exemplary embodiment of the application for judging whether a mold part passes the quality inspection based on a detection image. DETAILED DESCRIPTION

[0046] The embodiments of the application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the application, and cannot be understood as a limitation on the application.

[0047] As shown in an embodiment, Figure 1 The application provides a mold part quality control system, which comprises a camera module, an image processing module and a quality control module.

[0048] The camera module is used to obtain a detection image of a mold part.

[0049] The image processing module is used to judge whether the mold part passes the quality inspection based on the detection image.

[0050] The quality control module is used to count the number of mold parts that do not pass the quality inspection, obtain a statistical result, and perform quality control on the mold parts based on the statistical result.

[0051] As shown in the embodiment, Figure 2 Judging whether the mold part passes the quality inspection based on the detection image comprises:

[0052] Performing grayscale processing on the detection image to obtain a grayscale image.

[0053] The foreground image is obtained by foreground detection on the grayscale image.

[0054] After the noise reduction processing on the foreground image, it is judged whether the mold part passes the quality inspection.

[0055] In the quality inspection of the mold part by using machine vision, the foreground image corresponding to the detection image of the mold part is obtained first, then the noise reduction processing is performed on the foreground image, and then it is judged whether the mold part passes the quality inspection based on the result of the noise reduction processing, thereby effectively reducing the number of pixel points entering the noise reduction processing stage, improving the efficiency of the quality inspection of the mold part by using machine vision, and facilitating the quality inspection of the mold part.

[0056] Preferably, the camera module comprises a camera and a fill light;

[0057] The camera is used to obtain the detection image of the mold part;

[0058] The fill light is used to provide illuminating light for the camera.

[0059] Specifically, the fill light can provide illuminating light for the camera in a constant or flashing manner. The camera can be arranged above the device carrying the mold part, such as a conveyor belt, and obtain the detection image of the mold part in a vertical shooting manner.

[0060] Further, the number of cameras can be multiple, and the multiple cameras obtain the detection image of the mold part from different directions respectively.

[0061] Preferably, the image processing module comprises a storage unit, a grayscale unit, a foreground detection unit and an image recognition unit;

[0062] The storage unit is used to store the comparison image;

[0063] The grayscale unit is used to perform grayscale processing on the detection image and the comparison image respectively, and obtain the grayscale image and the grayscale comparison image;

[0064] The foreground detection unit is used to perform foreground detection on the grayscale image based on the grayscale comparison image, and obtain the foreground image;

[0065] The image recognition unit is used to judge whether the mold part passes the quality inspection after the noise reduction processing on the foreground image.

[0066] Preferably, the camera module is also used to obtain the comparison image.

[0067] Preferably, the comparison image is different from the shooting time of the detection image, and the comparison image is an image shot when the mold part is outside the shooting range of the camera module.

[0068] The contrast image does not contain the mold parts that need to be inspected, and other shooting conditions are the same except for the shooting time.

[0069] Preferably, the gray processing comprises:

[0070] For the image P, the gray processing is performed on P by using the following formula:

[0071] grayP(x,y)=w1×R(x,y)+w2×G(x,y)+w3×B(x,y)

[0072] In the formula, grayP represents the image obtained after the gray processing of the image P; grayP(x,y) represents the gray value of the pixel point with the coordinate (x,y) in grayP; R, G and B respectively represent the images corresponding to the red component, the green component and the blue component of the image P in the RGB color space; R(x,y), G(x,y) and B(x,y) respectively represent the pixel values of the pixel points with the coordinate (x,y) in R, G and B; w1, w2 and w3 represent the preset weighting coefficients.

[0073] Preferably, the foreground detection of the gray image based on the gray contrast image to obtain the foreground image comprises:

[0074] For the pixel point Pixel in the gray image, the judgment value of Pixel is calculated by using the following formula:

[0075] judval(Pixel)=grayimg(Pixel)-graycontimg(Pixel)|

[0076] In the formula, judval(Pixel) represents the judgment value of the pixel point Pixel, grayimg(Pixel) represents the gray value of the pixel point Pixel in the gray image grayimg, and graycontimg(Pixel) represents the gray value of the pixel point corresponding to Pixel in the gray contrast image;

[0077] If the judgment value is greater than the set judgment value threshold, the pixel point Pixel in the gray image is stored in the foreground set frtviwset.

[0078] The foreground image is obtained based on the pixel points in frtviwset.

[0079] Since the gray contrast image does not contain the pixel points corresponding to the mold parts, the pixel points with the changed pixel values can be obtained by comparing the pixel points with the same coordinates in the gray contrast image and the gray image, so that the pixel points constituting the foreground image are obtained based on the amplitude of the pixel point change.

[0080] Compared with the way of obtaining the foreground image by using the Otsu method, the watershed algorithm and the like, the speed is faster.

[0081] Preferably, the obtaining the foreground image based on the pixel points in the frtviwset comprises:

[0082] obtaining a minimum circumscribed rectangle Z of the pixel points in the frtviwset;

[0083] composing the foreground image by the pixel points in the minimum circumscribed rectangle Z range in the gray image.

[0084] Since the present application detects the change of the pixel value to perform the foreground detection, the result of the detection may appear the hole, therefore, the present application eliminates the hole by obtaining the minimum circumscribed rectangle, since the mold parts are originally continuous, therefore, the present application obtains the area of the mold parts more accurately.

[0085] Preferably, the judging whether the mold parts pass the quality inspection after the noise reduction processing of the foreground image comprises:

[0086] adopting the adaptive noise reduction algorithm to perform the noise reduction processing on the foreground image to obtain a noise reduction image;

[0087] performing the image segmentation processing on the noise reduction image to obtain a segmentation image;

[0088] obtaining the feature information of the segmentation image;

[0089] judging whether the mold parts pass the quality inspection based on the feature information.

[0090] Since the foreground image is obtained by the minimum circumscribed rectangle, therefore, the pixel points not belonging to the mold parts are also included in the foreground image, however, compared with the gray image, the proportion of the pixel points not belonging to the mold parts in the foreground image has been greatly reduced, at this time, the noise reduction processing can effectively reduce the pressure of the noise reduction processing and improve the speed of the noise reduction processing. After the noise reduction processing, the present application further removes the pixel points not belonging to the mold parts in the foreground image by the image segmentation method, so that the segmentation image does not contain the pixel points not belonging to the mold parts, which is beneficial to improve the accuracy of the feature information extracted subsequently.

[0091] Preferably, the adopting the adaptive noise reduction algorithm to perform the noise reduction processing on the foreground image to obtain a noise reduction image comprises:

[0092] performing the noise detection on the foreground image to obtain a set of noise pixel points setofnoipix;

[0093] respectively calculating the interference degree coefficient of each noise pixel point in the setofnoipix;

[0094] The noisy pixels are sorted from high to low according to the interference level coefficient to obtain a set sorset;

[0095] For the first noisy pixel noipix1 in the sorset, noise reduction is performed on noipix1 using the following method:

[0096] In the foreground image, obtain the set of pixels in the region to be denoised, which is a range of K×K pixels centered on noipix1;

[0097] An adaptive noise reduction algorithm is used to denoise the image imgtbdnoi1 composed of pixels in pixtbdnoi1 to obtain the denoised image afimgtbdnoi1.

[0098] The foreground image is updated using afimgtbdnoi1 to obtain image firupdimg1;

[0099] For the nth noisy pixel in the sorset, noipix n For n≥2, use the following method to process noipix n Noise reduction processing:

[0100] In image firupdimg n-1 Obtain from noipix n The set of pixels in the region to be denoised, centered on a point of interest and within a range of size K×K. n ;

[0101] An adaptive noise reduction algorithm is used for pixtbdnoi n The image composed of pixels in imgtbdnoi n Denoising is performed to obtain the denoised image afimgtbdnoi n ;

[0102] Use afimgtbdnoi n For image firupdimg n-1 Update and obtain image firupdimg n .

[0103] Existing noise reduction processing generally involves performing noise reduction on noisy pixels based on the same original image. During the processing, the original image, which serves as the reference, remains unchanged until the noise reduction results for all noisy pixels are calculated. Then, the pixel values ​​of the corresponding pixels in the original image are replaced based on the noise reduction results.

[0104] In the application, in addition to the first noise pixel point being based on the original foreground image, other noise pixel points are based on the updated image, which can effectively use the pixel value of the noise-reduced pixel point as a reference for noise reduction, thereby improving the accuracy of the noise reduction result of other pixel points.

[0105] The application first sorts the noise pixel points according to the interference degree coefficient. The greater the interference degree coefficient, the greater the influence of the noise pixel point on the surrounding pixel points. Therefore, it needs to be processed first, which facilitates subsequent processing of other noise pixel points and can obtain more accurate reference.

[0106] Preferably, the interference degree coefficient is calculated by the following formula:

[0107]

[0108] In the formula, intdegfac v represents the interference degree coefficient of the pixel point v, neipixset represents the set of 8-neighbor pixel points of the pixel point v, gray u and gray v respectively represent the pixel values of the pixel point u and the pixel point v.

[0109] The interference degree coefficient is mainly calculated from the sum of the pixel value difference between the noise pixel point and the surrounding pixel points. The greater the difference, the greater the interference degree coefficient.

[0110] Preferably, the foreground image is updated using afimgtbdnoi1 to obtain image firupdimg1, which includes:

[0111] Obtain the pixel point pix a in afimgtbdnoi1 n-1 ; b

[0112] Update the pixel value of the pixel point pix b to the pixel value of the pixel point pix a .

[0113] Preferably, the image firupdimg n is updated using afimgtbdnoi n-1 to obtain image firupdimg n , which includes:

[0114] Obtain the pixel point pix n in afimgtbdnoi c in firupdimg​n-1 corresponding to the pixel point pix d

[0115] the pixel value of the pixel point pix d is updated to the pixel value of the pixel point pix c .

[0116] Preferably, the adaptive noise reduction algorithm comprises:

[0117] For the image D, the process of using the adaptive noise reduction algorithm to perform noise reduction processing on the image D comprises:

[0118] wavelet decomposing the image D to obtain a high-frequency coefficient image hifrimg and a low-frequency coefficient image lofrimg;

[0119] For the high-frequency coefficient image hifrimg, noise reduction processing is performed in the following manner:

[0120] If hifrimg≤dorvalmi, the high-frequency coefficient image hifrimg is subjected to noise reduction processing using the following function:

[0121]

[0122] If dorvalmi<hifrimg<dorvalma, the high-frequency coefficient image hifrimg is subjected to noise reduction processing using the following function:

[0123]

[0124] ×noivstvar

[0125] If dorvalma≤hifrimg, the high-frequency coefficient image hifrimg is subjected to noise reduction processing using the following function:

[0126] ahifrimg=|hifrimg|

[0127] ​Wherein, dorvalmi and dorvalma are preset first comparison value and second comparison value, dorvalmi is less than dorvalma, ahifrimg represents an image obtained after noise reduction processing on hifrimg; thmr(hifrimg) represents an adaptive function, when hifrimg is less than 0, the value of thmr(hifrimg) is -delta, when hifrimg is 0, the value of thmr(hifrimg) is -delta, when hifrimg is greater than 0, the value of thmr(hifrimg) is delta, delta is an integer greater than 0, onc and twc represent two control parameters respectively, the value range of onc is [0.29, 0.39], the value range of twc is [0.79, 0.89],

[0128] Reconstructing ahifrimg and lofrimg, an image D after noise reduction is obtained.

[0129] In the present application, wavelet decomposition is first carried out, then the high-frequency coefficient image obtained is processed, and finally the processed high-frequency coefficient image and the low-frequency coefficient image are reconstructed to obtain accurate noise reduction results. In the process of noise reduction, two comparison values are set to select different noise reduction functions for the high-frequency coefficient image to adaptively reduce noise, and the degree of correspondence between the function and the high-frequency coefficient image is high, which is conducive to improving the accuracy of the noise reduction result.

[0130] Preferably, the noise-reduced image is subjected to image segmentation processing to obtain a segmented image, comprising:

[0131] The noise-reduced image is subjected to image segmentation processing using a region growing algorithm, and the pixel points in the noise-reduced image are divided into pixel points belonging to the mold parts and pixel points not belonging to the mold parts;

[0132] All pixel points belonging to the mold parts are combined to form a segmented image.

[0133] Preferably, the quality control module comprises a statistical unit, a calculation unit and a warning unit.

[0134] The characteristic information is input into a pre-trained neural network model to determine whether the mold parts pass the quality inspection.

[0135] The neural network model is mainly used to determine whether the characteristic information matches the various types of defects set, and if it matches, it means that the mold parts have corresponding types of defects and cannot pass the quality inspection. Specifically, the defects generally include burrs, cracks, etc.

[0136] Preferably, the quality control module comprises a statistical unit, a calculation unit and a warning unit.

[0137] The statistical unit is configured to count the number of mold parts of the same production batch that fail the quality inspection, and obtain a statistical result;

[0138] The calculation unit is configured to calculate a pass rate of the mold parts of the same production batch based on the statistical result;

[0139] The early warning unit is configured to send a warning prompt to the manager when the pass rate is less than a set pass rate threshold.

[0140] Specifically, the warning prompt can be in the form of a warning pop-up window popping up on the manager's office computer.

[0141] Preferably, the calculation of the pass rate of the mold parts of the same production batch based on the statistical result comprises:

[0142] calculating the pass rate of the mold parts;

[0143]

[0144] wherein pasrat represents the pass rate, num1 represents the number of mold parts of the same production batch that fail the quality inspection, and num2 represents the total number of mold parts of the same production batch.

[0145] Those skilled in the art will understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0146] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks

[0147] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects can be implemented on practitioners of ordinary skill in the art will appreciate that various modifications and changes can be made to the embodiments described above without departing from the scope of the present disclosure. Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects can be implemented on practitioners of ordinary skill in the art will appreciate that various modifications and changes can be made to the embodiments described above without departing from the scope of the present disclosure.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects can be implemented on practitioners of ordinary skill in the art will appreciate that various modifications and changes can be made to the embodiments described above without departing from the scope of the present disclosure. Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects can be implemented on practitioners of ordinary skill in the art will appreciate that various modifications and changes can be made to the embodiments described above without departing from the scope of the present disclosure.

[0149] The above description is made in connection with the preferred embodiments of the application. The application is not limited to the preferred embodiments, and various modifications or variations of the application can be made without departing from the scope of the application.

Claims

1. A quality control system for mold components, characterized in that, Includes a camera module, an image processing module, and a quality control module; The camera module is used to acquire inspection images of mold parts; The image processing module is used to determine whether mold parts have passed quality inspection based on the detected images; The quality control module is used to count the number of mold parts that fail quality inspection, obtain statistical results, and perform quality control on mold parts based on the statistical results. Among them, determining whether mold parts have passed quality inspection based on the detected images includes: The detected image is converted to grayscale to obtain a grayscale image; Perform foreground detection on a grayscale image to obtain a foreground image; After denoising the foreground image, the determination of whether the mold parts pass quality inspection includes: An adaptive noise reduction algorithm is used to denoise the foreground image to obtain a denoised image; The denoised image is segmented to obtain a segmented image; Obtain feature information from the segmented image; Determine whether mold parts have passed quality inspection based on feature information; The step of using an adaptive noise reduction algorithm to denoise the foreground image to obtain a denoised image includes: Noise detection is performed on the foreground image to obtain the set of noisy pixels, setofnoipix; Calculate the interference level coefficient for each noisy pixel in setofnoipix; The noisy pixels are sorted from high to low according to the interference level coefficient to obtain a set sorset; For the first noisy pixel noipix1 in the sorset, noise reduction is performed on noipix1 using the following method: In the foreground image, obtain the set of pixels in the region to be denoised, which is a range of K×K pixels centered on noipix1; An adaptive noise reduction algorithm is used to denoise the image imgtbdnoi1 composed of pixels in pixtbdnoi1 to obtain the denoised image afimgtbdnoi1. The foreground image is updated using afimgtbdnoi1 to obtain image firupdimg1; For the nth noisy pixel in the sorset, noipix n For n≥2, use the following method to process noipix n Noise reduction processing: In image firupdimg n-1 Obtain from noipix n The set of pixels in the region to be denoised, centered on a point of interest and within a range of size K×K. n ; An adaptive noise reduction algorithm is used for pixtbdnoi n The image composed of pixels in imgtbdnoi n Denoising is performed to obtain the denoised image afimgtbdnoi n ; Use afimgtbdnoi n For image firupdimg n-1 Update and obtain image firupdimg n .

2. The mold component quality control system according to claim 1, characterized in that, The camera module includes a camera and a fill light; The camera is used to acquire inspection images of mold parts; A fill light is used to provide illumination for the camera.

3. The mold component quality control system according to claim 1, characterized in that, The image processing module includes a storage unit, a grayscale unit, a foreground detection unit, and an image recognition unit; The storage unit is used to store comparison images; The grayscale unit is used to perform grayscale processing on the detection image and the comparison image respectively to obtain a grayscale image and a grayscale comparison image; The foreground detection unit is used to perform foreground detection on grayscale images based on grayscale contrast images to obtain foreground images; The image recognition unit is used to perform noise reduction processing on the foreground image to determine whether the mold parts have passed quality inspection.

4. A mold component quality control system according to claim 3, characterized in that, The comparison image and the detection image were captured at different times. The comparison image is an image captured when the mold parts are outside the shooting range of the camera module.

5. A mold component quality control system according to claim 3, characterized in that, The grayscale processing includes: For image P, the following formula is used to convert P to grayscale: grayP(x,y)=w1×R(x,y)+w2×G(x,y)+w3×B(x,y) In the formula, grayP represents the image obtained after grayscale processing of image P; grayP(x,y) represents the grayscale value of the pixel at coordinate (x,y) in grayP; R, G, and B represent the red, green, and blue components of image P in the RGB color space, respectively; R(x,y), G(x,y), and B(x,y) represent the pixel values ​​of the pixels at coordinate (x,y) in R, G, and B, respectively; and w1, w2, and w3 represent preset weighting coefficients.

6. A mold component quality control system according to claim 3, characterized in that, The method of performing foreground detection on a grayscale image based on a grayscale contrast image to obtain a foreground image includes: For a pixel in a grayscale image, the judgment value of the pixel is calculated using the following formula; judval(Pixel)=|grayimg(Pixel)-graycontimg(Pixel)| In the formula, judval(Pixel) represents the judgment value of pixel Pixel, grayimg(Pixel) represents the gray value of pixel Pixel in grayimg image grayimg, and graycontimg(Pixel) represents the gray value of the corresponding pixel in grayimg image. If the judgment value is greater than the set judgment value threshold, then the pixel in the grayscale image is stored in the foreground set frtviwset; Obtain the foreground image based on the pixels in frtviwset.

7. A mold component quality control system according to claim 1, characterized in that, The quality control module includes a statistical unit, a calculation unit, and an early warning unit; The statistical unit is used to count the number of mold parts that failed quality inspection in the same production batch and obtain statistical results; The calculation unit is used to calculate the pass rate of mold parts in the same production batch based on statistical results; The early warning unit is used to issue an early warning to management personnel when the pass rate is less than the set pass rate threshold.

8. A mold component quality control system according to claim 7, characterized in that, The calculation of the pass rate of mold parts in the same production batch based on statistical results includes: Calculate the pass rate of mold parts: In the formula, pasrat represents the pass rate, num1 represents the number of mold parts that failed quality inspection in the same production batch, and num2 represents the total number of mold parts in the same production batch.

Images