Humidity sensing film defect detection method and detection system

Abstract

The invention relates to the technical field of humidity sensing film defect detection, in particular to a humidity sensing film defect detection method and system. Constructing an image acquisition system, gradually and finely adjusting the shooting angles of the LED lamp and the camera, searching a shooting visual angle capable of highlighting the color difference between the transparent humidity sensing film and the carbon line to the greatest extent, and acquiring and obtaining an image; the image is preprocessed, the image is filtered in a bilateral filtering mode, and the image is enhanced; carrying out segmentation processing on the preprocessed image by adopting a Canny edge detection algorithm through an edge-based image segmentation method; and carrying out gray level detection on the image, carrying out image gray level analysis and color mapping, and identifying the coverage condition of the humidity sensing film in the image. According to the scheme, the coverage condition of the humidity sensing film in the image can be identified more accurately, and the performance of the hygristor element can be evaluated more comprehensively.

Description

Technical Field

[0001] The present invention relates to the technical field of moisture-sensitive film defect detection, and in particular to a moisture-sensitive film defect detection method and a detection system. Background Art

[0002] In the field of modern science and technology, humidity-sensitive resistors, as an important sensor device, have been widely used in many fields, such as environmental monitoring, medical diagnosis, industrial automation, and power battery production. As a key component of humidity-sensitive resistors, humidity-sensitive films play an important role in sensing humidity changes and converting them into electrical signals. However, humidity-sensitive films often face a series of preparation and quality control challenges during industrial preparation.

[0003] Defects in the humidity sensing film have a significant impact on the performance and stability of the humidity sensing resistor element. These defects may include structural unevenness, cracks, oxide residues, and other problems that may exist during the preparation process. These defects will cause inconsistent changes in the conductivity of the humidity sensing film, which in turn affects the sensitivity, response time, and reliability of the humidity sensing resistor element.

[0004] Therefore, surface defect detection plays a vital role in the parts production process as a "defender" to prevent defective products from entering the market and a "key man" to ensure the performance and reliability of humidity sensitive resistor components. Traditional defect detection methods, such as manual inspection, microscope observation and electrical testing, although simple and intuitive and can provide a certain degree of information, have certain limitations in terms of high precision and high efficiency, and the detection is easily affected by subjective factors, and the ability to detect subtle defects is limited.

[0005] Therefore, based on the above problems, there is an urgent need for an efficient, accurate and non-destructive method for detecting defects in humidity-sensitive films to improve the preparation process and quality of humidity-sensitive resistor elements, thereby ensuring the stability of their performance. Summary of the invention

[0006] The current humidity-sensitive film defect detection has certain limitations in terms of high precision and high efficiency. The present invention provides a humidity-sensitive film defect detection method and a detection system to solve the above problems.

[0007] To achieve the above object, the present invention is implemented through the following technical solutions: The present application discloses a method for detecting defects in a moisture-sensitive film, comprising: S1: Build an image acquisition system, gradually fine-tune the shooting angles of the LED light and the camera, find the shooting angle that can maximize the color difference between the transparent moisture-sensitive film and the carbon line, and collect and acquire the image; S2: Preprocessing the image, filtering the image by using bilateral filtering, and enhancing the image; the enhancement processing step includes: For a grayscale image of size m×n, at any position (x,y), where 1≤x≤m, 1≤y≤n, the corresponding pixel grayscale value is ,in , where L is the total number of gray levels of the image, and the probability of occurrence of different gray levels It can be expressed as: , , For the m×n grayscale image, after using histogram equalization enhancement, it can be expressed as follows: , Among them, the function Represents the mapping relationship between the gray level i of the image before enhancement and the gray level i′=f(i) after enhancement; S3: Using the edge-based image segmentation method, the Canny edge detection algorithm is used to segment the preprocessed image; S4: grayscale detection is performed on the image, grayscale analysis and color mapping are performed on the image, and the coverage of the moisture-sensitive film in the image is identified.

[0008] Adopting the above technical solution: The above solution can more accurately identify the coverage of the humidity-sensitive film in the image, and then make a more comprehensive evaluation of the performance of the humidity-sensitive resistor element. This research method not only improves the efficiency of analysis, but also provides strong support for the quality control and optimized design of humidity-sensitive resistor elements.

[0009] Optionally, the bilateral filtering includes: Get the spatial elements and pixel value elements, and multiply each pixel value by the spatial weight and the pixel value weight to get the new pixel value.

[0010] The above technical solution is adopted: bilateral filtering is selected as the filtering method for this study; image enhancement is performed through the adaptive histogram equalization method; in order to reduce image noise and improve image quality, the effects of various filtering methods are compared.

[0011] Optionally, the Canny edge detection algorithm includes: S31: Use Gaussian filter to reduce image noise; S32: Calculate the gradient direction and magnitude of each pixel in the image; S33: Use a non-maximum suppression algorithm to reduce the stray effect generated in the edge detection process, and determine whether the gradient strength of the current pixel is greater than that of the adjacent pixels in both the positive and negative gradient directions. If yes, identify it as an edge point, otherwise, identify it as a non-edge point and suppress it; S34: A dual threshold method is used to distinguish edge strengths. Pixels with gradient values ​​higher than the larger threshold are marked as strong edges; pixels with gradient values ​​between the two thresholds are marked as weak edges; pixels with gradient values ​​lower than the smaller threshold are ignored; S35: Eliminate isolated weak edges. If a weak edge has no strong edge in its 8-domain, delete it; otherwise, keep it.

[0012] Adopt the above technical solution: The above solution adopts the Canny edge detection algorithm, which has the characteristics of edge refinement and low error rate.

[0013] Optionally, the calculation formula of the Gaussian filter is: .

[0014] Optionally, the calculation formula for the gradient direction and amplitude of each pixel in the calculation image is: , .

[0015] Optionally, the color mapping adopts Inferno color mapping, and the Inferno color mapping can provide a gradient from black to red to bright yellow.

[0016] Adopting the above technical solution: The above solution adopts Inferno color mapping, which can help to accurately focus on the target object, thereby improving the accuracy and efficiency of recognition.

[0017] Optionally, the S4 also includes ROI extraction technology, which can identify and extract the region of interest in the image.

[0018] Adopt the above technical solution: In the above solution, ROI extraction is used to limit the scope of detection and focus attention on areas where defects may exist. By extracting the region of interest, defects or abnormal areas in the image can be effectively detected and located. Through accurate positioning and extraction, we can focus on the difference between the moisture-sensitive film and the carbon line, and perform defect detection classification.

[0019] Optionally, S4 includes: judging the coverage of the humidity sensitive film by comparing the grayscale curve of the defective element and the grayscale curve of the non-defective element; when the curve starts to rise steadily at a certain position and remains unchanged, the critical point position of the stable rise and maintenance is less than the set threshold, indicating that the humidity sensitive film is not completely covered; the critical point position of the stable rise and maintenance is greater than the set threshold, indicating that the humidity sensitive resistor element is a non-defective element.

[0020] Adopting the listed technical solution: the above solution can determine whether the humidity sensitive resistor element is a defective element.

[0021] A moisture-sensitive film defect detection system, applied to any one of the detection methods described above, is characterized by comprising: Image acquisition unit: The image acquisition unit includes an industrial camera, a lens, an optical arc slide, and an LED light. The best image acquisition effect can be achieved by adjusting the lens parameters and the positional relationship between the industrial camera, the optical arc slide, and the LED light. Image processing unit: The image processing unit is used to perform image denoising, enhancement, geometric correction and edge detection on the image collected by the image collection unit; Image analysis unit: The image analysis unit is used to extract the required features and make selections; Defect detection unit: used to detect defective parts of an image. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 This is a flow chart of the moisture-sensitive film defect detection method of the present application; Figure 2 This is the flow chart of the Canny edge detection algorithm for this application; Figure 3 For this application Figure 1 Step 4 of the flow chart; Figure 4 This is the grayscale curve of the humidity-sensitive film of this application; Figure 5 This is a comparison chart of the grayscale curve of this application; Figure 6 This is a flow chart of the fault testing method for this application. DETAILED DESCRIPTION

[0024] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0025] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0026] See also Figure 1-6 The present application embodiment discloses a method for detecting defects of a moisture-sensitive film, comprising: S1: Build an image acquisition system, gradually fine-tune the shooting angles of the LED light and the camera, find the shooting angle that can maximize the color difference between the transparent moisture-sensitive film and the carbon line, and collect and acquire the image; The image acquisition system constructed in the above scheme includes the selection of equipment and camera, such as resolution, frame rate, dynamic range, etc. In this embodiment, the SK9281 color CMOS industrial camera is preferably used; The choice of lens includes aperture, resolution and focusing capability. In this embodiment, the A9-1628-6MP area array lens is preferably used. This series has seven focal lengths, covering 6mm~50mm; this lens has a full field of view 6MP resolution and good edge clarity.

[0027] Regarding the choice of lighting, this system chooses to use a variety of light sources to form the best shooting lighting, including the use of point light source LED and ring light source LED distributed lighting and the use of the two light sources in combination.

[0028] In addition, the choice of lighting angle is also included. This design chooses oblique lighting. By adjusting the angle between the camera imaging center axis and the center line of the light source, the best reflection angle of the polymer film surface is found, thereby highlighting the surface details of the polymer film more clearly.

[0029] S2: preprocess the image, filter the image using bilateral filtering, and enhance the image; In traditional image processing, in order to enhance the contrast of an image, histogram equalization is usually used for image enhancement. This method redistributes the pixel values ​​of an image so that the histogram of the image is more evenly distributed within the brightness range, thereby improving the visual effect and visualization features of the image.

[0030] The core idea of ​​histogram equalization is to redistribute the pixel values ​​of the image so that they are evenly distributed within the new pixel range, thereby enhancing the less obvious details in the original image and improving the contrast of the overall image. For a grayscale image of size m×n, at any position (x, y), where 1≤x≤m, 1≤y≤n, the corresponding pixel grayscale value is ,in , where L is the total number of gray levels of the image, and the probability of occurrence of different gray levels It can be expressed as: , , For the m×n grayscale image, after enhancement using histogram equalization, it can be expressed as follows: , Among them, the function Represents the mapping relationship between the gray level i of the image before enhancement and the gray level i′=f(i) after enhancement; The implementation of histogram equalization is very simple, and in many cases it can effectively enhance the visual effect of the image. However, histogram equalization may cause the noise of the image to be amplified, and it cannot take into account the characteristics of the local structure of the image. Therefore, in some cases, it is necessary to combine other image enhancement techniques or use adaptive histogram equalization methods to process image areas with uneven grayscale distribution.

[0031] S3: Using the edge-based image segmentation method, the Canny edge detection algorithm is used to segment the preprocessed image; S4: grayscale detection is performed on the image, grayscale analysis and color mapping are performed on the image, and the coverage of the moisture-sensitive film in the image is identified.

[0032] The above scheme can more accurately identify the coverage of the humidity-sensitive film in the image, and then make a more comprehensive evaluation of the performance of the humidity-sensitive resistor element. This research method not only improves the efficiency of analysis, but also provides strong support for the quality control and optimized design of humidity-sensitive resistor elements.

[0033] like Figure 6As shown in , bilateral filtering is a commonly used nonlinear filter used for smoothing filtering and edge preservation in image processing. Compared with traditional linear filters, bilateral filters can better smooth images while retaining edge details. The basic principle of bilateral filtering is that for each pixel in the filtering window, the bilateral filter must consider both spatial factors and pixel value factors, and multiply each pixel value by the spatial weight and the pixel value weight to obtain a new pixel value. The bilateral filter smoothes the image without destroying the details and edges of the image. The bilateral filtering includes: Get the spatial elements and pixel value elements, and multiply each pixel value by the spatial weight and pixel value weight to get the new pixel value. Select bilateral filtering as the filtering method for this study; use the adaptive histogram equalization method to enhance the image; reduce image noise and improve image quality, and compare the effects of various filtering methods.

[0034] The Canny edge detection algorithm includes: S31: Use Gaussian filter to reduce image noise; S32: Calculate the gradient direction and magnitude of each pixel in the image; S33: Use a non-maximum suppression algorithm to reduce the stray effect generated in the edge detection process, and determine whether the gradient strength of the current pixel is greater than that of the adjacent pixels in both the positive and negative gradient directions. If yes, identify it as an edge point, otherwise, identify it as a non-edge point and suppress it; S34: A double threshold method is used to distinguish edge strengths. Pixels with gradient values ​​higher than the larger threshold are marked as strong edges; pixels with gradient values ​​between the two thresholds are marked as weak edges; pixels with gradient values ​​lower than the smaller threshold are ignored; S35: Eliminate isolated weak edges. If a weak edge has no strong edge in its 8-domain, delete it; otherwise, keep it.

[0035] The above scheme adopts the Canny edge detection algorithm, which has the characteristics of edge refinement and low error rate.

[0036] The calculation formula of the Gaussian filter is: .

[0037] The calculation formula for the gradient direction and amplitude of each pixel in the calculation image is: , .

[0038] The color mapping adopts Inferno color mapping, which shows a significant difference when presenting the foreground and background. Under Inferno color mapping, the color of irrelevant information such as the background presents a relatively uniform characteristic, which prevents the background information from interfering with the recognition of the detected object in the foreground. This feature is particularly important in image processing and object recognition because it can help to accurately focus on the target object, thereby improving the accuracy and efficiency of recognition. The Inferno color mapping can provide a gradient from black to red to bright yellow. The above scheme adopts Inferno color mapping, which can help to accurately focus on the target object, thereby improving the accuracy and efficiency of recognition.

[0039] The S4 also includes ROI extraction technology, which can identify and extract the region of interest in the image. In the above scheme, ROI extraction is used to limit the scope of detection and focus attention on the area where defects may exist. By extracting the region of interest, defects or abnormal areas in the image can be effectively detected and located. Through accurate positioning and extraction, the difference between the moisture-sensitive film and the carbon line can be focused on, and defect detection classification can be performed.

[0040] like Figure 4 and Figure 5 As shown in the figure, an example is used to illustrate that after a series of image processing is performed on the pictures collected by the image acquisition system, the grayscale value of the processed humidity-sensitive film image is studied, and the grayscale value curve is drawn. It can be found that the peaks represent the white gaps between the carbon lines and the carbon line areas not covered by the humidity-sensitive film, and the troughs correspond to the carbon line areas covered by the humidity-sensitive film. The grayscale value curve has 6 peaks, representing 6 white gaps, with an average width of 6.67, and the grayscale value reached by the peaks is about 200; there are 6 troughs, representing 6 vertical carbon lines, and the grayscale value reached by the troughs is about 140. The grayscale value of the position where the curve changes from low to high and remains unchanged is about 210.

[0041] By comparing the grayscale curve of the defective element and the grayscale curve of the non-defective element, the coverage of the humidity sensitive film can be judged. When the curve starts to rise steadily at a certain position and remains unchanged, the critical point position of the stable rise and the continuous rise is less than the set threshold, which means that the humidity sensitive film is not completely covered. The critical point position of the stable rise and the continuous rise is greater than the set threshold, which means that the humidity sensitive resistor element is a non-defective element. The above scheme can realize the judgment of whether the humidity sensitive resistor element is a defective element.

[0042] A moisture-sensitive film defect detection system, applied to any of the above detection methods, comprising: Image acquisition unit: The image acquisition unit includes an industrial camera, a lens, an optical arc slide, and an LED light. The best image acquisition effect can be achieved by adjusting the lens parameters and the positional relationship between the industrial camera, the optical arc slide, and the LED light. Image processing unit: The image processing unit is used to perform image denoising, enhancement, geometric correction and edge detection on the image collected by the image collection unit; Image analysis unit: The image analysis unit is used to extract the required features and make selections; Defect detection unit: used to detect defective parts of an image.

[0043] Unless otherwise specified, the device components involved in the above embodiments are all conventional device components, and the connection methods and control methods involved are all conventional connection methods and control methods unless otherwise specified.

[0044] The present invention has been described in detail above in conjunction with the embodiments. However, those skilled in the art will appreciate that, without departing from the spirit of the present invention, the specific parameters in the above embodiments may be changed to form a plurality of specific embodiments, which are all within the common variation range of the present invention and will not be described in detail herein.

Claims

1. A method for detecting defects in a moisture-sensitive film, characterized in that: include: S1: Build an image acquisition system, gradually fine-tune the shooting angles of the LED light and the camera, find the shooting angle that can maximize the color difference between the transparent moisture-sensitive film and the carbon line, and collect and acquire the image; S2: preprocess the image, filter the image using bilateral filtering, and enhance the image; The enhancement processing step comprises: For a grayscale image of size m×n, at any position (x,y), where 1≤x≤m, 1≤y≤n, the corresponding pixel grayscale value is ,in , where L is the total number of gray levels of the image, and the probability of occurrence of different gray levels It can be expressed as: ， ， For the m×n grayscale image, after enhancement using histogram equalization, it can be expressed as follows: ， Among them, the function Represents the mapping relationship between the gray level i of the image before enhancement and the gray level i′=f(i) after enhancement; S3: Using the edge-based image segmentation method, the Canny edge detection algorithm is used to segment the preprocessed image; S4: grayscale detection is performed on the image, grayscale analysis and color mapping are performed on the image, and the coverage of the moisture-sensitive film in the image is identified.

2. A moisture-sensitive film defect detection method according to claim 1, characterized in that: The bilateral filtering includes: Get the spatial elements and pixel value elements, and multiply each pixel value by the spatial weight and the pixel value weight to get the new pixel value.

3. A moisture-sensitive film defect detection method according to claim 1, characterized in that: Said include: S31: Use Gaussian filter to reduce image noise; S32: Calculate the gradient direction and magnitude of each pixel in the image; S33: Use a non-maximum suppression algorithm to reduce the stray effect generated in the edge detection process, and determine whether the gradient strength of the current pixel is greater than that of the adjacent pixels in both the positive and negative gradient directions. If yes, identify it as an edge point, otherwise, identify it as a non-edge point and suppress it; S34: A dual threshold method is used to distinguish edge strengths. Pixels with gradient values ​​higher than the larger threshold are marked as strong edges; pixels with gradient values ​​between the two thresholds are marked as weak edges; pixels with gradient values ​​lower than the smaller threshold are ignored; S35: Eliminate isolated weak edges. If a weak edge has no strong edge in its 8-domain, delete it; otherwise, keep it.

4. A moisture-sensitive film defect detection method according to claim 3, characterized in that: The calculation formula of the Gaussian filter is: 。 5. A moisture-sensitive film defect detection method according to claim 3, characterized in that: The calculation formula for the gradient direction and amplitude of each pixel in the calculation image is: ， 。 6. A moisture-sensitive film defect detection method according to claim 1, characterized in that: The color mapping adopts Inferno color mapping, and the Inferno color mapping can provide a gradient from black to red to bright yellow.

7. A moisture-sensitive film defect detection method according to claim 1, characterized in that: The S4 also includes ROI extraction technology, which can identify and extract the region of interest in the image.

8. A moisture-sensitive film defect detection method according to claim 1, characterized in that: The S4 includes: judging the coverage of the humidity sensitive film by comparing the grayscale curve of the defective element and the grayscale curve of the non-defective element; when the curve starts to rise steadily at a certain position and remains unchanged, the position of the critical point of the stable rise and the continuous rise is less than the set threshold, which means that the humidity sensitive film is not completely covered; the position of the critical point of the stable rise and the continuous rise is greater than the set threshold, which means that the humidity sensitive resistor element is a non-defective element.

9. A moisture-sensitive film defect detection system, applied to the detection method according to any one of claims 1 to 8, characterized in that: include: Image acquisition unit: The image acquisition unit includes an industrial camera, a lens, an optical arc slide, and an LED light. The best image acquisition effect can be achieved by adjusting the lens parameters and the positional relationship between the industrial camera, the optical arc slide, and the LED light. Image processing unit: The image processing unit is used to perform image denoising, enhancement, geometric correction and edge detection on the image collected by the image collection unit; Image analysis unit: The image analysis unit is used to extract the required features and make selections; Defect detection unit: used to detect defective parts of an image.

Images