Rainbow film intelligent quality detection method and system based on image recognition

Abstract

The invention discloses a rainbow film intelligent quality detection method and system based on image recognition, and relates to the technical field of image processing, and the method comprises the steps: carrying out the convolution of a rainbow film image through employing a Sobel operator, generating a binary mask, generating a filtered morphological brightness image, carrying out the one-dimensional Fourier transform of the filtered morphological brightness image, and obtaining a morphological brightness image; using an arc tangent binary function to calculate a direction angle of a complex rotation, extracting a main phase field, calculating a continuous phase field, performing convolution operation on the continuous phase field, and generating a smoothed continuous phase field; and calculating a gradient mode of the smoothed continuous phase field to generate a binary defect mask. According to the method, a high-quality form brightness graph is generated through multi-level feature extraction and self-adaptive segmentation, stripe continuity and defect distinguishability are enhanced, a defect area is self-adaptively recognized by constructing a smooth and continuous phase field in combination with a statistical threshold, and the robust recognition capability of a complex interference pattern is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a rainbow film intelligent quality detection method and system based on image recognition. BACKGROUND

[0002] With the rapid development of image processing and pattern recognition technology, image recognition has been widely used in industrial detection, medical diagnosis, security monitoring and other fields. In the field of precision manufacturing, the analysis of optical interference images has become an important means of product quality detection. In the products of optical thin film, coated film material, rainbow film and other complex interference structures, the image-based detection method can identify subtle defects and structural abnormalities in a non-contact, high-precision and high-efficiency manner.

[0003] The existing image recognition detection method for rainbow film pattern still has certain limitations. The rainbow film image presents strong interference fringe structure, and its spatial frequency distribution is complex. The existing method has problems such as edge blur and false structure misjudgment in the non-continuous area of the fringe, especially when processing high-frequency or curvature discontinuous area. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a rainbow film intelligent quality detection method and system based on image recognition, which solves the problem that the rainbow film image presents strong interference fringe structure, and its spatial frequency distribution is complex. The existing method has problems such as edge blur and false structure misjudgment in the non-continuous area of the fringe, especially when processing high-frequency or curvature discontinuous area.

[0006] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a rainbow film intelligent quality detection method based on image recognition, which includes the following steps: Collecting a rainbow film image and performing preprocessing, using a Sobel operator to convolve the rainbow film image, using a nonlinear transformation to calculate a curvature potential function, calculating an adaptive threshold, generating a binary mask, using median filtering to fill in the non-continuous area, and generating a morphological intensity map; Converting the morphological intensity map from the spatial domain to the frequency domain, calculating the gradient direction angle of the morphological intensity map, constructing a high-pass filter mask, performing a point multiplication operation on the frequency domain representation and the high-pass filter mask, converting the filtered frequency domain representation from the frequency domain back to the spatial domain to generate a filtered morphological intensity map, performing one-dimensional Fourier transform on the filtered morphological intensity map along the x direction, calculating the main period, calculating the period difference field, using the inverse tangent binary function to calculate the direction angle of the complex curl, extracting the main phase field, calculating the continuous phase field, performing convolution operation on the continuous phase field, and generating the smoothed continuous phase field; The gradient modulus of the smoothed continuous phase field is calculated, a statistical threshold is used to set a detection threshold, a binary defect mask is generated, and a visual interface is constructed to display the binary defect mask.

[0007] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the Sobel operator is used to convolve the iris image to generate a morphological intensity map, including: The Sobel operator is used to convolve the iris image to calculate the gradient of the iris image, the normal curvature is calculated based on the gradient of the iris image, the second derivative is calculated using the Laplace operator, the curvature potential function is calculated using a nonlinear transformation, and normalization processing is performed. An adaptive threshold is calculated using the median statistical method to generate a binary mask, and the median filter is used to fill in the discontinuous area equal to 0 in the binary mask to generate a morphological intensity map.

[0008] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the filtered morphological intensity map is generated, including: The two-dimensional fast Fourier transform is used to convert the morphological intensity map from the spatial domain to the frequency domain to obtain a frequency domain representation. The Sobel operator is used to calculate the gradient of the morphological intensity map, the gradient direction angle of the morphological intensity map is calculated, the histogram of the gradient direction angle is counted, and the main direction angle is determined to construct a high-pass filter mask. The point multiplication operation is performed on the frequency domain representation and the high-pass filter mask to obtain the filtered frequency domain representation.

[0009] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the continuous phase field is calculated, the continuous phase field is convolved to generate a smoothed continuous phase field, including: The filtered morphological intensity map is subjected to one-dimensional Fourier transform along the x direction to calculate the main period. The filtered morphological intensity map is subjected to period difference along the x and y directions to construct a difference field and calculate the period difference field. Based on the period difference field, a complex map is constructed, and the complex helicity is calculated by partial derivative, the direction angle of the complex helicity is calculated using the arctangent binary function, and the main phase field is extracted. The continuous phase field and the integer compensation term are initialized, the center point of the filtered morphological intensity map is taken as the starting point, the integer compensation term is initialized to 0, the path integral method is used to start from the starting point, the breadth-first search is used to traverse all pixels, the phase difference is calculated, the integer compensation term is updated, and the continuous phase field is calculated. The offset relative to the center pixel is set using a fixed window indexing method, a Gaussian kernel is defined, a convolution operation is performed on the continuous phase field, and a smoothed continuous phase field is generated.

[0010] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the gradient modulus of the smoothed continuous phase field is calculated, and a binary defect mask is generated. The gradient modulus of the smoothed continuous phase field is calculated. A detection threshold is set using statistical threshold segmentation, and a binary defect mask is generated.

[0011] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the visualization interface is constructed to display the binary defect mask, which includes: The binary defect mask is visualized using the front-end framework React.js. The user after real-name verification is allowed to view.

[0012] As a preferred scheme of the image recognition-based iris intelligent quality detection method, the collection of the iris image and the preprocessing thereof includes: An industrial-grade camera is used to collect the iris image, and denoising and normalization processing are performed.

[0013] In a second aspect, the present application provides an image recognition-based iris intelligent quality detection system, which includes: A morphology module is used to collect the iris image and perform preprocessing, a Sobel operator is used to convolve the iris image, a non-linear transformation is used to calculate a curvature potential function, an adaptive threshold is calculated, a binary mask is generated, a median filter is used to fill in non-continuous areas, and a morphological intensity map is generated. A filter smoothing module is used to convert the morphological intensity map from the spatial domain to the frequency domain, calculate the gradient direction angle of the morphological intensity map, construct a high-pass filter mask, perform a point multiplication operation on the frequency domain representation and the high-pass filter mask, convert the filtered frequency domain representation from the frequency domain back to the spatial domain, generate a filtered morphological intensity map, perform one-dimensional Fourier transform on the filtered morphological intensity map along the x direction, calculate the main period, calculate the period difference field, use the inverse tangent binary function to calculate the direction angle of the complex rotational velocity, extract the main phase field, calculate the continuous phase field, perform a convolution operation on the continuous phase field, and generate a smoothed continuous phase field. A detection visualization module is used to calculate the gradient modulus of the smoothed continuous phase field, set a detection threshold using statistical threshold segmentation, generate a binary defect mask, and construct a visualization interface to display the binary defect mask.

[0014] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the computer program, when executed by the processor, implements any step of the image recognition based iris smart quality detection method according to the first aspect of the present application.

[0015] In a fourth aspect, the present application provides a computer readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any step of the image recognition based iris smart quality detection method according to the first aspect of the present application.

[0016] The present application has the advantages that: the present application generates a high-quality morphological intensity map through multi-level feature extraction and adaptive segmentation, enhances the continuity of the fringes and the distinguishability of the defects, constructs a smooth and continuous phase field, and combines with a statistical threshold to adaptively identify the defect regions, thereby improving the robust recognition capability for complex interference patterns. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 The flowchart of the image recognition based iris smart quality detection method in embodiment 1.

[0019] Figure 2 The schematic diagram of the image recognition based iris smart quality detection system in embodiment 1. DETAILED DESCRIPTION

[0020] In order to make the above objectives, features and advantages of the present application more apparent and understandable, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0021] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced without the specific details, other than those described herein, and it is understood that the present application is not limited to the embodiments described herein and that the embodiments can be practiced with or without the specific details. Embodiments of the present application, therefore, can include numerous alternatives, modifications, and equivalents.

[0022] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments.

[0023] Embodiment 1, Reference Figure 1 As a first embodiment of the present application, the embodiment provides an image recognition-based iris smart quality detection method, comprising the following steps: S1, collecting iris images and performing preprocessing, using Sobel operator to convolve the iris images, using nonlinear transformation to calculate curvature potential function, calculating adaptive threshold, generating binary mask, using median filter to fill in non-continuous area, and generating morphological brightness graph; Specifically, collecting iris images and performing preprocessing, comprising: Collecting iris images using an industrial-grade camera and performing denoising and normalization processing.

[0024] Industrial-grade cameras are used for image acquisition to ensure image clarity and stability; through gray scale normalization processing, the interference of environmental brightness changes on feature recognition is suppressed, and the universality and robustness of the processing algorithm are improved.

[0025] Further, using Sobel operator to convolve the iris images to generate morphological brightness graph, comprising: Using Sobel operator to convolve the iris images, calculating the gradient of the iris images, and calculating the normal curvature based on the gradient of the iris images, the formula is: , , , , wherein is the normalized iris image, x and y are the horizontal and vertical coordinates of the image, and the center pixel position is specified, and are the gradients of the iris image, representing the gradients along the x and y directions, and are the horizontal and vertical convolution kernels of the Sobel operator, and n is the gradient normal direction, is a small constant of the normal curvature to prevent division by zero of the gradient amplitude, is the normal curvature; Using Laplace operator to calculate second derivative, the formula is: , wherein is the second derivative, representing the result of the Laplace operator, is the neighborhood set of the Laplace operator, using fixed neighborhood method to set, is the number of neighborhood points, i and j are the horizontal and vertical direction offsets, respectively, representing the pixel coordinates relative to The coordinate offset; The curvature potential function is calculated using a nonlinear transformation and then normalized. The formula is as follows: , in Let be the curvature potential function, which characterizes the geometric continuity and curvature response of the fringes; An adaptive threshold is calculated using median statistics to generate a binary mask, as shown in the formula: , , in For adaptive threshold, The threshold offset was set using an experimental optimization method to enhance the ability to distinguish discontinuous regions. For a binary mask, 0 represents a discontinuous pseudo-structure region and 1 represents a continuous stripe region; Median filtering is used to fill in discontinuous regions equal to 0 in the binary mask, generating a morpholuminance map.

[0026] The Sobel operator has the advantages of strong directionality and good edge preservation, making it suitable for extracting weak and complex interference fringe structures in rainbow film images. Normal curvature is used to characterize the degree of local surface changes in the image, effectively identifying pseudo-fringe regions (discontinuous curvature) and real fringe regions (smooth curvature). The Laplacian operator is sensitive to edge response and can accurately locate the positions of brightness abrupt changes and texture changes in the image. Combined with the first derivative characteristics for fusion processing, it improves the recognition rate of discontinuous pseudo-structure regions. Through curvature and second derivative co-modeling, it realizes multi-scale expression of image structure. The normalization process unifies the regional response scale and reduces the interference of external environmental factors. The adaptive threshold is generated based on local statistical features, avoiding the risk of missegmentation caused by fixed thresholds. Median filtering can smoothly fill small broken regions without destroying the real fringe structure, restoring the morphological integrity of the fringe. It has good repair capabilities for random noise or small breaks in the image.

[0027] S2. Transform the morpholuminance map from the spatial domain to the frequency domain, calculate the gradient direction angle of the morpholuminance map, construct a high-pass filter mask, perform a dot product operation on the frequency domain representation and the high-pass filter mask, transform the filtered frequency domain representation back to the spatial domain, generate the filtered morpholuminance map, perform a one-dimensional Fourier transform on the filtered morpholuminance map along the x-direction, calculate the principal period, calculate the period difference field, use the arctangent binary function to calculate the direction angle of the complex curl, extract the principal phase field, calculate the continuous phase field, perform a convolution operation on the continuous phase field, and generate a smoothed continuous phase field. Specifically, the generated filtered morpholuminance map includes: The morphological intensity map is converted from spatial domain to frequency domain using two-dimensional fast Fourier transform to obtain a frequency domain representation; The morphological intensity map gradient is calculated using a Sobel operator, and a morphological intensity map gradient direction angle is calculated, according to the formula: , , , wherein and are the morphological intensity map gradient, is the morphological intensity map, is the gradient direction angle, is a small constant of the gradient direction angle, is a two-dimensional convolution operation; A histogram of the gradient direction angle is counted to determine a main direction angle, according to the formula: , wherein is the main direction angle, corresponding to the histogram peak value, representing the main direction of the interference fringes, is the histogram of the gradient direction angle; A high-pass filter mask is constructed, according to the formula: , , , wherein is the frequency domain frequency direction angle, u and v being the horizontal frequency coordinate and the vertical frequency coordinate in the frequency domain respectively, is the frequency amplitude, representing the distance of the frequency component to the origin, is the frequency radius threshold, set using a main frequency peak positioning method, is the half-width of the main direction angle range, set using a direction distribution analysis method, is the high-pass filter mask, in binary form, is a small constant of the frequency domain frequency direction angle; The frequency domain representation and the high-pass filter mask are point-multiplied to retain the main direction high-frequency component, to obtain a filtered frequency domain representation, according to the formula: , wherein is the filtered frequency domain representation, is the frequency domain representation; The filtered frequency domain representation is converted from the frequency domain back to the spatial domain using two-dimensional inverse fast Fourier transform, to generate a filtered morphological intensity map.

[0028] The period structure such as stripe is converted into clear frequency component, the background noise is stripped, the main frequency direction is extracted and high-pass filtered, the main direction of stripe is accurately positioned, the subsequent directional filtering is ensured to be targeted, the multi-directional overlapping interference is overcome, the main shaft of structure is purified, the high-frequency stripe information in the main direction is accurately reserved, the background interference in other directions is suppressed, the direction consistency and structure integrity of image are effectively enhanced, the filtered image has higher stripe clarity and direction consistency, and a clean background and main structure highlight foundation are laid for main period extraction and phase reconstruction.

[0029] Further, a continuous phase field is calculated, and a convolution operation is performed on the continuous phase field to generate a smoothed continuous phase field, including: The filtered morphological brightness image is subjected to one-dimensional Fourier transform along the x direction to calculate the main period, and the formula is: , , wherein fmax is the frequency corresponding to the maximum amplitude, is the maximum amplitude, is the height of the filtered morphological brightness image, and is is a down rounding operation, is the filtered morphological brightness image, is one-dimensional fast Fourier transform along the x direction, and outputs a complex frequency spectrum, f is a frequency, and T is the main period; The filtered morphological brightness image is subjected to period difference along the x and y directions to construct a difference field, and the period difference field is calculated, and the formula is: , , , wherein and are period differences along the x and y directions, respectively, and represent intensity changes after the main period is offset, is the period difference field, is a small constant of the period difference field; Based on the period difference field, a complex image is constructed, and a complex curl is calculated through partial derivative, and the formula is: , , wherein is the complex image, and s is an imaginary unit, is a complex exponential function, and are partial derivatives of the complex image along the x and y directions, respectively, and are calculated using a Sobel operator, The complex curl is a vector field representing the rotational property of the phase field. The direction angle of the complex curl is calculated using the arctangent binary function to extract the principal phase field, and the formula is: , wherein is the principal phase field, representing the phase distribution of the interference fringes, is the imaginary part of the complex curl, is the real part of the complex curl; Initialize the continuous phase field and the integer compensation term, and set the center point of the filtered morphological intensity map as the starting point. The integer compensation term is initialized to 0. Using the path integral method, the starting point is used to traverse all pixels using the breadth-first search. The phase difference is calculated, the integer compensation term is updated, and the continuous phase field is calculated, and the formula is: , , , wherein and are the principal phase field values, and are the neighborhood pixel coordinates, which are set using the fixed window neighborhood method, is the continuous phase field, and are the integer compensation terms, which are integer values, is the phase difference value; The offset relative to the center pixel is set using the fixed window index method, and the Gaussian kernel is defined. The continuous phase field is convolved to generate the smoothed continuous phase field, and the formula is: , , wherein is the Gaussian kernel, m and h are the offsets relative to the center pixel, is the standard deviation of the Gaussian distribution, which is set using the data-driven method to control the smoothing degree, and k is the window half-width, which is set using the window size experience setting method, is the smoothed continuous phase field.

[0030] The periodic difference field is used to determine the relative offset trend of the periodic structure, the direction consistency of the periodic phase field is enhanced, the periodic difference is introduced into the complex space modeling, the complex rotation provides the phase offset and the rotation characteristic expression of the image structure, the stability of the phase reconstruction is improved, the complex modeling is combined with the path integral, the continuous phase field with high consistency and strong anti-interference ability is constructed, the Gaussian smoothing control scale is used to suppress noise while preserving the true structure, the burr effect caused by local phase discontinuity is effectively eliminated, and the robustness of the phase data for subsequent three-dimensional reconstruction, contour identification and other tasks is improved.

[0031] S3, calculating the gradient modulus of the smoothed continuous phase field, using statistical threshold segmentation to set a detection threshold, generating a binary defect mask, and constructing a visual interface to display the binary defect mask; Specifically, the gradient modulus of the smoothed continuous phase field is calculated, and the binary defect mask is generated, comprising: The gradient modulus of the smoothed continuous phase field is calculated, and the formula is: , Wherein and are the partial derivatives of the smoothed continuous phase field along the x and y directions, respectively, and are calculated using the Sobel operator, is the gradient modulus. The statistical threshold segmentation is used to set a detection threshold, the area higher than the threshold is marked as a potential defect area, and a binary defect mask is generated, and the formula is: , , Wherein is the detection threshold, and are the global mean and standard deviation of the gradient modulus, respectively, is the binary defect mask.

[0032] The defect area is often accompanied by a sharp phase jump, which is represented as a high amplitude response in the gradient domain. The background area with slow phase change has a lower response in the gradient modulus, while the edge or abnormal area is clearly highlighted. Based on the adaptive threshold of the image itself statistical distribution, the over-detection or missed detection problem caused by manual parameter setting is avoided. Since the statistical quantity is calculated from the global gradient distribution, the local mutation area is significantly enhanced and the background has less impact.

[0033] Further, the visual interface is constructed to display the binary defect mask, comprising: The visual interface is constructed using the front-end framework React.js, and the binary defect mask is visualized. Allowing the user to check after real-name verification.

[0034] The present application provides a clear and intuitive defect display view, which facilitates the user to quickly judge the defect nature, position and shape, controls the data access permission through real-name verification, and meets the privacy management requirements of sensitive images of scientific research institutions or industrial systems.

[0035] Embodiment 2, refer to Figure 2 For the second embodiment of the present application, an image recognition-based iris intelligent quality detection system comprises: A morphology collection module is configured to collect and pre-process the iris image, convolve the iris image using a Sobel operator, calculate a curvature potential function using a nonlinear transformation, calculate an adaptive threshold, generate a binary mask, fill in non-continuous regions using median filtering, and generate a morphological intensity map. A filter smoothing module is configured to convert the morphological intensity map from a spatial domain to a frequency domain, calculate a morphological intensity map gradient direction angle, construct a high-pass filter mask, perform a point multiplication operation on the frequency domain representation and the high-pass filter mask, convert the filtered frequency domain representation from the frequency domain back to the spatial domain, generate a filtered morphological intensity map, perform one-dimensional Fourier transform on the filtered morphological intensity map along the x direction, calculate a main period, calculate a period difference field, calculate the direction angle of the complex curl using an inverse tangent binary function, extract a main phase field, calculate a continuous phase field, perform a convolution operation on the continuous phase field, and generate a smoothed continuous phase field. A detection visualization module is configured to calculate the gradient modulus of the smoothed continuous phase field, set a detection threshold using a statistical threshold segmentation, generate a binary defect mask, and construct a visualization interface to display the binary defect mask.

[0036] The present embodiment also provides a computer device suitable for the image recognition-based iris intelligent quality detection method, comprising a memory and a processor; the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to realize the image recognition-based iris intelligent quality detection method as proposed in the above embodiments.

[0037] The computer device can be a terminal, which includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved by WIFI, operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0038] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the method for intelligent iris quality detection based on image recognition as proposed in the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk or an optical disk.

[0039] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A method for intelligent quality detection of rainbow film based on image recognition, characterized in that: Includes the following steps: Collect rainbow film images and preprocess them. Convolve the rainbow film images using the Sobel operator, calculate the curvature potential function using nonlinear transformation, calculate the adaptive threshold, generate a binary mask, fill discontinuous regions using median filtering, and generate a morpholuminance map. The morpholuminance map is transformed from the spatial domain to the frequency domain. The gradient direction angle of the morpholuminance map is calculated, a high-pass filter mask is constructed, and a dot product operation is performed on the frequency domain representation and the high-pass filter mask. The filtered frequency domain representation is transformed back from the frequency domain to the spatial domain to generate a filtered morpholuminance map. The filtered morpholuminance map is then subjected to a one-dimensional Fourier transform along the x-direction to calculate the principal period. The period difference field is calculated, and the direction angle of the complex curl is calculated using the arctangent binary function. The principal phase field is extracted, and the continuous phase field is calculated. The continuous phase field is then convolved to generate a smoothed continuous phase field. The gradient magnitude of the smoothed continuous phase field is calculated, the detection threshold is set using statistical threshold segmentation, a binary defect mask is generated, and a visualization interface is constructed to display the binary defect mask.

2. The intelligent quality detection method for rainbow film based on image recognition as described in claim 1, characterized in that: The step of convolving the rainbow film image using the Sobel operator to generate a morphological brightness map includes: The Sobel operator is used to convolve the iris image, the gradient of the iris image is calculated, the normal curvature is calculated based on the gradient of the iris image, the second derivative is calculated using the Laplacian operator, the curvature potential function is calculated using nonlinear transformation, and then normalization is performed. An adaptive threshold is calculated using median statistics to generate a binary mask. Median filtering is then used to fill in discontinuous regions equal to 0 in the binary mask, generating a morpholuminance map.

3. The intelligent quality detection method for rainbow film based on image recognition as described in claim 2, characterized in that: The generated filtered morphology-luminance map includes: The morpholuminance map is transformed from the spatial domain to the frequency domain using a two-dimensional fast Fourier transform to obtain a frequency domain representation; The Sobel operator is used to calculate the gradient of the morphology brightness map, the gradient direction angle of the morphology brightness map is calculated, the histogram of the gradient direction angle is statistically analyzed, the main direction angle is determined, and a high-pass filter mask is constructed. Perform a dot product operation on the frequency domain representation and the high-pass filter mask to obtain the filtered frequency domain representation.

4. The intelligent quality detection method for rainbow film based on image recognition as described in claim 3, characterized in that: The calculation of the continuous phase field, which involves performing a convolution operation on the continuous phase field to generate a smoothed continuous phase field, includes: Perform a one-dimensional Fourier transform on the filtered morpholuminance map along the x-direction to calculate the main period; The filtered morpholuminance map is used to construct a difference field by periodically differencing along the x and y directions, and the periodic difference field is calculated. Based on the periodic difference field, a complex graph is constructed, and the complex curl is calculated by partial derivatives. The direction angle of the complex curl is calculated by the arctangent bivariate function, and the principal phase field is extracted. Initialize the continuous phase field and integer compensation term. Take the center point of the filtered morpholuminance map as the starting point, initialize the integer compensation term to 0, use the path integral method to start from the starting point, use breadth-first search to traverse all pixels, calculate the phase difference, update the integer compensation term, and calculate the continuous phase field. Using a fixed window indexing method to set the offset relative to the center pixel, a Gaussian kernel is defined, and a convolution operation is performed on the continuous phase field to generate a smoothed continuous phase field.

5. The intelligent quality detection method for rainbow film based on image recognition as described in claim 4, characterized in that: The gradient magnitude of the smoothed continuous phase field is calculated to generate a binary defect mask, including: Calculate the gradient magnitude of the smoothed continuous phase field; Statistical thresholding is used to set the detection threshold and generate a binary defect mask.

6. The intelligent quality detection method for rainbow film based on image recognition as described in claim 5, characterized in that: The construction of a visual interface to display the binary defect mask includes: A visual interface is built using the front-end framework React.js to visualize the binary defect mask; Users who have completed real-name verification are allowed to view this information.

7. The intelligent quality detection method for rainbow film based on image recognition as described in claim 6, characterized in that: The process of collecting and preprocessing rainbow film images includes: Images of the rainbow film were collected using an industrial-grade camera and then denoised and normalized.

8. A rainbow film intelligent quality inspection system based on image recognition, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The morphology collection module is used to collect rainbow film images and perform preprocessing. It uses the Sobel operator to convolve the rainbow film images, uses nonlinear transformation to calculate the curvature potential function, calculates the adaptive threshold, generates a binary mask, uses median filtering to fill discontinuous regions, and generates a morphology brightness map. The filtering and smoothing module is used to transform the morpholuminescence map from the spatial domain to the frequency domain, calculate the gradient direction angle of the morpholuminescence map, construct a high-pass filter mask, perform a dot product operation on the frequency domain representation and the high-pass filter mask, transform the filtered frequency domain representation back from the frequency domain to the spatial domain, generate a filtered morpholuminescence map, perform a one-dimensional Fourier transform on the filtered morpholuminescence map along the x-direction, calculate the principal period, calculate the period difference field, use the arctangent binary function to calculate the direction angle of the complex curl, extract the principal phase field, calculate the continuous phase field, perform a convolution operation on the continuous phase field, and generate a smoothed continuous phase field. The detection visualization module is used to calculate the gradient magnitude of the smoothed continuous phase field, set the detection threshold using statistical threshold segmentation, generate a binary defect mask, and build a visualization interface to display the binary defect mask.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the image recognition-based intelligent quality detection method for rainbow film as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the image recognition-based intelligent quality detection method for rainbow film as described in any one of claims 1 to 7.

Images