Image defect traceability detection method and system suitable for mobile phone touch screen

Abstract

The invention relates to the field of mobile phone image analysis, and discloses an image defect traceability detection method and system suitable for a mobile phone touch screen, and the method comprises the following steps: positioning abnormal positions of pixel points of a sample image in a mobile phone, and carrying out the defect type classification of the abnormal positions of the pixel points of the sample image in combination with a support vector machine model, and finally, defect traceability detection and defect solution design are carried out on different classification results. According to the invention, pixel analysis can be carried out on a sample image on a mobile phone, defect traceability analysis of mobile phone hardware and software and a shooting process can be realized based on an analysis result, a solution design is realized, and the purpose of preventing and solving an image display problem is realized.

Description

Technical Field

[0001] The present invention relates to the field of mobile phone image analysis, and in particular to an image defect tracing detection method and system suitable for a mobile phone touch screen. Background Art

[0002] Blurred images in mobile phones may involve problems with the mobile phone itself, or they may be related to the mobile phone touch screen (i.e., display screen). For example, tiny stains such as dust, fingerprints, and grease may adhere to the mobile phone lens, affecting the clarity of photos or videos. At the same time, camera damage, lens instability, or other hardware problems may also cause blurred images. And there are problems with the image. If the mobile phone touch screen itself is damaged or aged, such as defects in the LCD, failure of internal components (such as graphics cards), etc., the image will be blurred. Displaying a still image for a long time will cause permanent image traces on the screen, which is called "burn-in". The burn-in phenomenon will also cause the displayed image to be blurred. In summary, the causes of blurred images in mobile phones may involve multiple aspects, including mobile phone cameras, mobile phone touch screens, shooting techniques, and environmental conditions. Therefore, an image defect tracing detection method and system suitable for mobile phone touch screens are proposed to determine the defect source of mobile phone images and achieve the purpose of defect repair based on the symptoms. Summary of the invention

[0003] The present invention overcomes the deficiencies of the prior art and provides an image defect tracing detection method and system suitable for a mobile phone touch screen.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is: A first aspect of the present invention provides an image defect tracing detection method applicable to a mobile phone touch screen, comprising the following steps: An image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image; For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis results; In the image defect analysis module, defect tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image.

[0005] Furthermore, in a preferred embodiment of the present invention, the image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image, specifically: Determine the target mobile phone, and calibrate the touch screen of the target mobile phone as the touch screen of the target mobile phone; Installing an image defect analysis module in the target mobile phone, wherein the image defect analysis module is a module that can determine the type of image defects and perform traceability detection on image defects, and determining a sample image in the target mobile phone; The sample image is an image used to test whether the target mobile phone and the touch screen of the target mobile phone have defects, and the target mobile phone and the touch screen of the target mobile phone are defect traceable according to the defect type on the sample image; Importing and mapping the sample image into the image defect analysis module, based on the image defect analysis module, recording the pixel value of each pixel point on the sample image, calibrating it to the actual pixel value, and determining the standard pixel value threshold of each pixel point on the sample image, calibrating it to the standard pixel value threshold; Pixel value analysis is performed on all pixels on the sample image, and pixels whose actual pixel values ​​are not within the standard pixel value threshold are marked as abnormal pixels. At the same time, the abnormal pixels are marked and highlighted on the sample image.

[0006] Furthermore, in a preferred embodiment of the present invention, for the abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis result, specifically: The support vector machine model is introduced into the image defect analysis module, and the threshold extraction method is used to perform threshold extraction processing on abnormal pixels in the sample image to obtain an abnormal pixel image; In the abnormal pixel image, a pixel grayscale co-occurrence matrix is ​​constructed, and based on the grayscale co-occurrence matrix of the abnormal pixel, the texture features of the abnormal pixel image are extracted, and at the same time, the abnormal pixel image is subjected to Fourier transform to obtain the frequency domain features of the abnormal pixel image, and finally, the color features of the pixels on the abnormal pixel image are extracted to obtain the color features of the abnormal pixel image; Combining the texture features, frequency domain features, and color features of the abnormal pixel point image into a feature vector, constructing a training data set based on the feature vector, and importing the training data set into the support vector machine model; A big data network is introduced, and all possible defect types of the sample image are retrieved based on the big data network, and all corresponding support vector machine model kernel functions are retrieved based on all possible defect types of the sample image, and calibrated as target kernel functions; Based on the objective and function, the training data set is cross-validated in the support vector machine model to implement data training of the support vector machine model, and the defect type existing in the abnormal pixel image is output after cross-validation and marked as the actual defect type of the image; The actual image defect types include sample image color defects, blur defects and image quality defects.

[0007] Furthermore, in a preferred embodiment of the present invention, in the image defect analysis module, defect source tracing detection is performed based on the defect type of the sample image, the source of the defect type of the sample image is determined, and a defect solution is designed in combination with the source of the defect type of the sample image, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a color defect, the standard color temperature parameters and standard color parameters of the abnormal pixel image are determined in the target mobile phone, and combined to form the standard color parameters, and the actual color parameters of the abnormal pixel image are determined at the same time; Calculate the parameter difference between the standard color parameter and the actual color parameter, preset the qualified parameter difference, perform a mobile phone system version detection on the target mobile phone, determine whether there is an available update version for the target mobile phone, and if so, install the available update version on the target mobile phone, and determine whether the parameter difference between the standard color parameter and the actual color parameter is not greater than the qualified parameter difference after installation; If yes, the source of the color defect of the sample image is calibrated as the target mobile phone system cause. If no, it is determined whether the color parameter adjustment in the target mobile phone can make the parameter difference between the standard color parameter and the actual color parameter not greater than the qualified parameter difference. If still not, then the color defect source of the sample image is calibrated as the hardware reason of the target mobile phone touch screen, the position of the abnormal pixel image on the target mobile phone touch screen is determined, and calibrated as the color abnormality position of the target mobile phone touch screen, and the hardware corresponding to the color abnormality position of the target mobile phone touch screen in the target mobile phone touch screen is determined, and calibrated as a type of abnormal hardware, and based on the big data network, all solutions for hardware maintenance of a type of abnormal hardware are retrieved; In the image defect analysis module, defect tracing detection is performed on sample images whose actual defect type is fuzzy defect, and defect solutions are designed.

[0008] Furthermore, in a preferred embodiment of the present invention, in the image defect analysis module, defect tracing detection is performed on sample images whose actual defect type is a fuzzy defect, and a defect solution is designed, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a blur defect, a gyroscope module of the target mobile phone is obtained, and the gyroscope module of the target mobile phone is connected to the image analysis module; After the gyroscope module of the target mobile phone is connected to the image analysis module, the recording timestamp of the sample image is determined, and based on the recording timestamp of the sample image, the jitter amplitude and jitter frequency of the gyroscope when the sample image is recorded are determined, and calibrated as the actual jitter amplitude and the actual jitter frequency of the gyroscope; Based on the big data network, the jitter amplitude and jitter frequency of the gyroscope corresponding to the sample image without blur defects are determined, and calibrated as the standard jitter amplitude and standard jitter frequency of the gyroscope; If the actual jitter amplitude and the actual jitter frequency of the gyroscope are greater than the corresponding standard jitter amplitude and the standard jitter frequency of the gyroscope, the blur defect source of the sample image is calibrated as the cause of shooting jitter; In the image defect analysis module, a blind deconvolution algorithm is introduced. An initial blur kernel is constructed based on the blind deconvolution algorithm, and a blur estimation value of the initial blur kernel is defined. The sample image is convolved with the blur estimation value of the initial blur kernel to obtain a clear estimate of the sample image. Based on the blind deconvolution algorithm, the sample image is iteratively convolved, and the maximum number of iterations is preset. If the number of iterations is equal to the maximum number of iterations, the iterative convolution of the sample image is stopped, and the sample image after iterative convolution is output and calibrated as the blind deconvolution sample image; Wherein, the blind deconvolution sample image does not have blur defects; If the actual jitter amplitude and actual jitter frequency of the gyroscope are not greater than the corresponding standard jitter amplitude and standard jitter frequency of the gyroscope, perform secondary fuzzy defect traceability on the target mobile phone and the target mobile phone touch screen, and design a defect solution based on the secondary traceability results.

[0009] Furthermore, in a preferred embodiment of the present invention, the target mobile phone and the target mobile phone touch screen are subjected to secondary fuzzy defect tracing, and a defect solution is designed based on the secondary tracing result, specifically: If the actual gyroscope jitter amplitude and the actual gyroscope jitter frequency are not greater than the corresponding gyroscope standard jitter amplitude and the gyroscope standard jitter frequency, a hardware damage analysis is performed on the camera of the target mobile phone to determine whether the camera of the target mobile phone has a hardware damage problem; If so, the blur defect source of the sample image is calibrated as the hardware cause of the target mobile phone camera, and the repair plan of the damaged hardware of the target mobile phone camera is retrieved and output in the big data network; If not, then update the camera application cache and the camera application in the target mobile phone. If the sample image does not have blur defects after the camera application cache and the camera application are updated, then the source of the blur defects in the sample image is calibrated as the target mobile phone camera software cause; If the sample image still has blur defects after the camera application cache is updated and the camera application is updated, the source of the blur defect of the sample image is calibrated as a hardware problem of the target mobile phone touch screen, and the target mobile phone touch screen is replaced to ensure that the sample image does not have blur defects.

[0010] The second aspect of the present invention further provides an image defect tracing detection system applicable to a mobile phone touch screen, the image defect tracing detection system comprising a memory and a processor, the memory storing an image defect tracing detection method, and when the image defect tracing detection method is executed by the processor, the following steps are implemented: An image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image; For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis results; In the image defect analysis module, defect tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image.

[0011] The present invention solves the technical defects existing in the background technology, and the present invention has the following beneficial effects: locating the abnormal position of the pixel point of the sample image in the mobile phone, and combining the support vector machine model to classify the defect type of the abnormal position of the pixel point of the sample image, and finally performing defect tracing detection and defect solution design on different classification results. The present invention can perform pixel analysis on the sample image on the mobile phone, and realize defect tracing analysis of the mobile phone hardware and software and the shooting process based on the analysis results, and realize solution design, so as to achieve the purpose of preventing and solving image display problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0013] Figure 1 A flow chart showing an image defect tracing detection method applicable to a mobile phone touch screen is shown; Figure 2 A flow chart of a method for defect tracing detection based on the defect type of a sample image is shown; Figure 3 A program view of an image defect tracing detection system suitable for a mobile phone touch screen is shown. DETAILED DESCRIPTION

[0014] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0015] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0016] Figure 1 A flow chart of an image defect tracing detection method applicable to a mobile phone touch screen is shown, comprising the following steps: S102: installing an image defect analysis module in the mobile phone, and marking pixels of the sample images with abnormalities by performing image analysis on the sample images; S104: For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and defect type classification of the sample image is performed based on the feature analysis result; S106: In the image defect analysis module, defect source tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image.

[0017] Furthermore, in a preferred embodiment of the present invention, the image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image, specifically: Determine the target mobile phone, and calibrate the touch screen of the target mobile phone as the touch screen of the target mobile phone; Installing an image defect analysis module in the target mobile phone, wherein the image defect analysis module is a module that can determine the type of image defects and perform traceability detection on image defects, and determining a sample image in the target mobile phone; The sample image is an image used to test whether the target mobile phone and the touch screen of the target mobile phone have defects, and the target mobile phone and the touch screen of the target mobile phone are defect traceable according to the defect type on the sample image; Importing and mapping the sample image into the image defect analysis module, based on the image defect analysis module, recording the pixel value of each pixel point on the sample image, calibrating it to the actual pixel value, and determining the standard pixel value threshold of each pixel point on the sample image, calibrating it to the standard pixel value threshold; Pixel value analysis is performed on all pixels on the sample image, and pixels whose actual pixel values ​​are not within the standard pixel value threshold are marked as abnormal pixels. At the same time, the abnormal pixels are marked and highlighted on the sample image.

[0018] It should be noted that the purpose of installing the image defect analysis module is to directly trace the source of defects of problematic images in the mobile phone and correct them, saving time. The sample image is an image that is retained in the mobile phone through various methods such as shooting, uploading, and storage on the target mobile phone. The image may have defects, such as blur defects, clarity defects, and other problems. In the image, not all positions are abnormal, and defects may only exist in some positions, so the image is analyzed by pixel points and pixel values, and the position corresponding to the pixel point with abnormal pixel value is the defect position. The purpose of highlighting abnormal pixels is to more clearly know the location coordinates of the abnormal position.

[0019] Furthermore, in a preferred embodiment of the present invention, for the abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis result, specifically: The support vector machine model is introduced into the image defect analysis module, and the threshold extraction method is used to perform threshold extraction processing on abnormal pixels in the sample image to obtain an abnormal pixel image; In the abnormal pixel image, a pixel grayscale co-occurrence matrix is ​​constructed, and based on the grayscale co-occurrence matrix of the abnormal pixel, the texture features of the abnormal pixel image are extracted, and at the same time, the abnormal pixel image is subjected to Fourier transform to obtain the frequency domain features of the abnormal pixel image, and finally, the color features of the pixels on the abnormal pixel image are extracted to obtain the color features of the abnormal pixel image; Combining the texture features, frequency domain features, and color features of the abnormal pixel point image into a feature vector, constructing a training data set based on the feature vector, and importing the training data set into the support vector machine model; A big data network is introduced, and all possible defect types of the sample image are retrieved based on the big data network, and all corresponding support vector machine model kernel functions are retrieved based on all possible defect types of the sample image, and calibrated as target kernel functions; Based on the objective and function, the training data set is cross-validated in the support vector machine model to implement data training of the support vector machine model, and the defect type existing in the abnormal pixel image is output after cross-validation and marked as the actual defect type of the image; The actual image defect types include sample image color defects, blur defects and image quality defects.

[0020] It should be noted that the defect types include image color defects, blur defects and image quality defects, but directly observing the image with the naked eye cannot accurately determine the defect type of the image, and it is necessary to combine the algorithm for detailed division. The algorithm needs to use a support vector machine model, wherein the support vector machine model is a multi-step algorithm for feature analysis and defect type classification of abnormal pixels in the image, also known as the SVM algorithm. First, the image is threshold segmented to remove the image positions where there are no abnormalities, simplify the calculation steps, and improve the efficiency of defect tracing. Secondly, texture feature extraction and frequency domain feature extraction are performed. The texture feature further describes the abnormal position in the image, while the frequency domain feature is the necessary feature data for training the SVM model, which can be obtained through Fourier transform. Construct a training data set, and train the support vector machine model based on the training data set. When training the support vector machine model, it is necessary to select a suitable kernel function, including a linear kernel function, a polynomial kernel function, etc. The cross-validation method includes a grid search method, etc., which is used to optimize the parameters of the SVM model to achieve the purpose of image defect type classification, and finally obtain the actual defect type of the image.

[0021] Figure 2 A flow chart of a method for defect tracing detection based on the defect type of a sample image is shown, comprising the following steps: S202: In the image defect analysis module, defect source tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image; S204: In the image defect analysis module, defect tracing detection is performed on the sample image whose actual defect type is a fuzzy defect, and a defect solution is designed; S206: Perform secondary fuzzy defect tracing on the target mobile phone and the target mobile phone touch screen, and design a defect solution based on the secondary tracing result.

[0022] Furthermore, in a preferred embodiment of the present invention, in the image defect analysis module, defect source tracing detection is performed based on the defect type of the sample image, the source of the defect type of the sample image is determined, and a defect solution is designed in combination with the source of the defect type of the sample image, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a color defect, the standard color temperature parameters and standard color parameters of the abnormal pixel image are determined in the target mobile phone, and combined to form the standard color parameters, and the actual color parameters of the abnormal pixel image are determined at the same time; Calculate the parameter difference between the standard color parameter and the actual color parameter, preset the qualified parameter difference, perform a mobile phone system version detection on the target mobile phone, determine whether there is an available update version for the target mobile phone, and if so, install the available update version on the target mobile phone, and determine whether the parameter difference between the standard color parameter and the actual color parameter is not greater than the qualified parameter difference after installation; If yes, the source of the color defect of the sample image is calibrated as the target mobile phone system cause. If no, it is determined whether the color parameter adjustment in the target mobile phone can make the parameter difference between the standard color parameter and the actual color parameter not greater than the qualified parameter difference. If still not, then calibrate the color defect source of the sample image as the hardware reason of the target mobile phone touch screen, determine the position of the abnormal pixel image on the target mobile phone touch screen, calibrate it as the color abnormality position of the target mobile phone touch screen, and determine the hardware corresponding to the color abnormality position of the target mobile phone touch screen in the target mobile phone touch screen, calibrate it as a type of abnormal hardware, and retrieve all solutions for hardware maintenance of a type of abnormal hardware based on the big data network.

[0023] It should be noted that we first analyze whether the sample image has color defects. By analyzing the color parameters of the lossless version of the sample image, that is, the standard color parameters and the actual color parameters, we can get the parameter difference and compare it with the standard parameter difference to determine whether there is a color defect. Because mobile phones are different, mobile phones with better performance may display colors closer to reality, but if the parameter difference is large, it is not a performance problem of the mobile phone, but other problems may occur directly on the mobile phone. First, determine the version of the mobile phone. An old version may affect the authenticity of the color display of the mobile phone. Secondly, analyze the touch screen of the mobile phone to determine the abnormal position of the sample image corresponding to the touch screen of the mobile phone. This position is the position of the hardware abnormality, and retrieve all solutions for hardware maintenance of a type of abnormal hardware through the big data network.

[0024] Furthermore, in a preferred embodiment of the present invention, in the image defect analysis module, defect tracing detection is performed on sample images whose actual defect type is a fuzzy defect, and a defect solution is designed, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a blur defect, a gyroscope module of the target mobile phone is obtained, and the gyroscope module of the target mobile phone is connected to the image analysis module; After the gyroscope module of the target mobile phone is connected to the image analysis module, the recording timestamp of the sample image is determined, and based on the recording timestamp of the sample image, the jitter amplitude and jitter frequency of the gyroscope when the sample image is recorded are determined, and calibrated as the actual jitter amplitude and the actual jitter frequency of the gyroscope; Based on the big data network, the jitter amplitude and jitter frequency of the gyroscope corresponding to the sample image without blur defects are determined, and calibrated as the standard jitter amplitude and standard jitter frequency of the gyroscope; If the actual jitter amplitude and the actual jitter frequency of the gyroscope are greater than the corresponding standard jitter amplitude and the standard jitter frequency of the gyroscope, the blur defect source of the sample image is calibrated as the cause of shooting jitter; In the image defect analysis module, a blind deconvolution algorithm is introduced. An initial blur kernel is constructed based on the blind deconvolution algorithm, and a blur estimation value of the initial blur kernel is defined. The sample image is convolved with the blur estimation value of the initial blur kernel to obtain a clear estimate of the sample image. Based on the blind deconvolution algorithm, the sample image is iteratively convolved, and the maximum number of iterations is preset. If the number of iterations is equal to the maximum number of iterations, the iterative convolution of the sample image is stopped, and the sample image after iterative convolution is output and calibrated as the blind deconvolution sample image; Wherein, the blind deconvolution sample image does not have blur defects; If the actual jitter amplitude and actual jitter frequency of the gyroscope are not greater than the corresponding standard jitter amplitude and standard jitter frequency of the gyroscope, perform secondary fuzzy defect traceability on the target mobile phone and the target mobile phone touch screen, and design a defect solution based on the secondary traceability results.

[0025] It should be noted that there is no color defect, and the blur defect analysis of the sample image is required. The blur defect of the sample image may be a problem of the mobile phone itself, or it may be a problem of shaking during shooting. First, analyze whether it is a shaking problem, and how to correct the blur of the picture caused by shaking. Get the gyroscope module of the mobile phone. The gyroscope can record the mobile phone's internal modules that affect the shooting clarity, such as the shaking frequency, rate, and amplitude of the mobile phone. The gyroscope records the jitter parameters when shooting the sample image, that is, the actual jitter parameters, and retrieves the standard jitter parameters for comparison. If the actual value is greater than the standard value, the source of the blur defect of the sample image is calibrated as the cause of shooting jitter, because the actual jitter amplitude and frequency are high, resulting in image blur. The sample image obtained under the cause of shooting jitter can be used to eliminate image motion blur through the blind deconvolution algorithm. Blind deconvolution is a method of estimating blur kernels and clear images through iterative optimization. It assumes that the blurred image is obtained by convolution operation of the clear image and blur kernel. First, initialize an estimated value of the blur kernel, and then use the estimated value and the blurred image to perform deconvolution operation to obtain an estimated value of the clear image. Next, the estimated value of the blur kernel is updated according to the difference between the estimated value of the clear image and the original blurred image. Through multiple iterations, the estimated values ​​of the blur kernel and the clear image are continuously optimized until the convergence condition is reached, that is, the number of iterations is equal to the maximum number of iterations, and finally the blind deconvolution sample image is obtained.

[0026] Furthermore, in a preferred embodiment of the present invention, the target mobile phone and the target mobile phone touch screen are subjected to secondary fuzzy defect tracing, and a defect solution is designed based on the secondary tracing result, specifically: If the actual gyroscope jitter amplitude and the actual gyroscope jitter frequency are not greater than the corresponding gyroscope standard jitter amplitude and the gyroscope standard jitter frequency, a hardware damage analysis is performed on the camera of the target mobile phone to determine whether the camera of the target mobile phone has a hardware damage problem; If so, the blur defect source of the sample image is calibrated as the hardware cause of the target mobile phone camera, and the repair plan of the damaged hardware of the target mobile phone camera is retrieved and output in the big data network; If not, then update the camera application cache and the camera application in the target mobile phone. If the sample image does not have blur defects after the camera application cache and the camera application are updated, then the source of the blur defects in the sample image is calibrated as the target mobile phone camera software cause; If the sample image still has blur defects after the camera application cache is updated and the camera application is updated, the source of the blur defect of the sample image is calibrated as a hardware problem of the target mobile phone touch screen, and the target mobile phone touch screen is replaced to ensure that the sample image does not have blur defects.

[0027] It should be noted that the actual jitter amplitude and actual jitter frequency of the gyroscope are not greater than the corresponding standard jitter amplitude and standard jitter frequency of the gyroscope, which proves that the blur defect of the sample image has nothing to do with the shooting, but is related to the mobile phone hardware. The mobile phone camera hardware may be damaged or cracked, or the internal application cache of the mobile phone may be too high, resulting in insufficient memory when the image is output, resulting in blurred image. It may also be a touch screen hardware problem, and the touch screen needs to be replaced directly.

[0028] like Figure 3 As shown, the second aspect of the present invention further provides an image defect tracing detection system suitable for a mobile phone touch screen, the image defect tracing detection system comprising a memory 31 and a processor 32, the memory 31 stores an image defect tracing detection method, and when the image defect tracing detection method is executed by the processor 32, the following steps are implemented: An image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image; For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis results; In the image defect analysis module, defect tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image.

[0029] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An image defect tracing detection method applicable to a mobile phone touch screen, characterized in that: The following steps are involved: An image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image; For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis results; In the image defect analysis module, defect tracing detection is performed based on the defect type of the sample image to determine the source of the defect type of the sample image, and a defect solution is designed in combination with the source of the defect type of the sample image.

2. The image defect tracing detection method for mobile phone touch screen according to claim 1, characterized in that: The image defect analysis module is installed in the mobile phone, and the pixel points of the sample image with abnormalities are marked by performing image analysis on the sample image, specifically: Determine the target mobile phone, and calibrate the touch screen of the target mobile phone as the touch screen of the target mobile phone; Installing an image defect analysis module in the target mobile phone, wherein the image defect analysis module is a module that can determine the type of image defects and perform traceability detection on image defects, and determining a sample image in the target mobile phone; The sample image is an image used to test whether the target mobile phone and the touch screen of the target mobile phone have defects, and the target mobile phone and the touch screen of the target mobile phone are defect traceable according to the defect type on the sample image; Importing and mapping the sample image into the image defect analysis module, based on the image defect analysis module, recording the pixel value of each pixel point on the sample image, calibrating it to the actual pixel value, and determining the standard pixel value threshold of each pixel point on the sample image, calibrating it to the standard pixel value threshold; Pixel value analysis is performed on all pixels on the sample image, and pixels whose actual pixel values ​​are not within the standard pixel value threshold are marked as abnormal pixels. At the same time, the abnormal pixels are marked and highlighted on the sample image.

3. The image defect tracing detection method applicable to mobile phone touch screens according to claim 1, characterized in that: For abnormal pixels, a support vector machine model is introduced into the image defect analysis module to perform feature analysis, and the defect type of the sample image is classified based on the feature analysis results, specifically: The support vector machine model is introduced into the image defect analysis module, and the threshold extraction method is used to perform threshold extraction processing on abnormal pixels in the sample image to obtain an abnormal pixel image; In the abnormal pixel image, a pixel grayscale co-occurrence matrix is ​​constructed, and based on the grayscale co-occurrence matrix of the abnormal pixel, the texture features of the abnormal pixel image are extracted, and at the same time, the abnormal pixel image is subjected to Fourier transform to obtain the frequency domain features of the abnormal pixel image, and finally, the color features of the pixels on the abnormal pixel image are extracted to obtain the color features of the abnormal pixel image; Combining the texture features, frequency domain features, and color features of the abnormal pixel point image into a feature vector, constructing a training data set based on the feature vector, and importing the training data set into the support vector machine model; A big data network is introduced, and all possible defect types of the sample image are retrieved based on the big data network, and all corresponding support vector machine model kernel functions are retrieved based on all possible defect types of the sample image, and calibrated as target kernel functions; Based on the objective and function, the training data set is cross-validated in the support vector machine model to implement data training of the support vector machine model, and the defect type existing in the abnormal pixel image is output after cross-validation and marked as the actual defect type of the image; The actual image defect types include sample image color defects, blur defects and image quality defects.

4. The image defect tracing detection method for a mobile phone touch screen according to claim 1, characterized in that: In the image defect analysis module, defect tracing detection is performed based on the defect type of the sample image, the source of the defect type of the sample image is determined, and a defect solution is designed in combination with the source of the defect type of the sample image, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a color defect, the standard color temperature parameters and standard color parameters of the abnormal pixel image are determined in the target mobile phone, and combined to form the standard color parameters, and the actual color parameters of the abnormal pixel image are determined at the same time; Calculate the parameter difference between the standard color parameter and the actual color parameter, preset the qualified parameter difference, perform a mobile phone system version detection on the target mobile phone, determine whether there is an available update version for the target mobile phone, and if so, install the available update version on the target mobile phone, and determine whether the parameter difference between the standard color parameter and the actual color parameter is not greater than the qualified parameter difference after installation; If yes, the source of the color defect of the sample image is calibrated as the target mobile phone system cause. If no, it is determined whether the color parameter adjustment in the target mobile phone can make the parameter difference between the standard color parameter and the actual color parameter not greater than the qualified parameter difference. If still not, then the color defect source of the sample image is calibrated as the hardware reason of the target mobile phone touch screen, the position of the abnormal pixel image on the target mobile phone touch screen is determined, and calibrated as the color abnormality position of the target mobile phone touch screen, and the hardware corresponding to the color abnormality position of the target mobile phone touch screen in the target mobile phone touch screen is determined, and calibrated as a type of abnormal hardware, and based on the big data network, all solutions for hardware maintenance of a type of abnormal hardware are retrieved; In the image defect analysis module, defect tracing detection is performed on sample images whose actual defect type is fuzzy defect, and defect solutions are designed.

5. The image defect tracing detection method applicable to mobile phone touch screens according to claim 4, characterized in that: In the image defect analysis module, defect tracing detection is performed on sample images whose actual defect type is blur defect, and defect solutions are designed, specifically: In the image defect analysis module, if the actual image defect type of the sample image is a blur defect, a gyroscope module of the target mobile phone is obtained, and the gyroscope module of the target mobile phone is connected to the image analysis module; After the gyroscope module of the target mobile phone is connected to the image analysis module, the recording timestamp of the sample image is determined, and based on the recording timestamp of the sample image, the jitter amplitude and jitter frequency of the gyroscope when the sample image is recorded are determined, and calibrated as the actual jitter amplitude and the actual jitter frequency of the gyroscope; Based on the big data network, the jitter amplitude and jitter frequency of the gyroscope corresponding to the sample image without blur defects are determined, and calibrated as the standard jitter amplitude and standard jitter frequency of the gyroscope; If the actual jitter amplitude and the actual jitter frequency of the gyroscope are greater than the corresponding standard jitter amplitude and the standard jitter frequency of the gyroscope, the blur defect source of the sample image is calibrated as the cause of shooting jitter; In the image defect analysis module, a blind deconvolution algorithm is introduced. An initial blur kernel is constructed based on the blind deconvolution algorithm, and a blur estimation value of the initial blur kernel is defined. The sample image is convolved with the blur estimation value of the initial blur kernel to obtain a clear estimate of the sample image. Based on the blind deconvolution algorithm, the sample image is iteratively convolved, and the maximum number of iterations is preset. If the number of iterations is equal to the maximum number of iterations, the iterative convolution of the sample image is stopped, and the sample image after iterative convolution is output and calibrated as the blind deconvolution sample image; Wherein, the blind deconvolution sample image does not have blur defects; If the actual jitter amplitude and actual jitter frequency of the gyroscope are not greater than the corresponding standard jitter amplitude and standard jitter frequency of the gyroscope, perform secondary fuzzy defect traceability on the target mobile phone and the target mobile phone touch screen, and design a defect solution based on the secondary traceability results.

6. The image defect tracing detection method applicable to a mobile phone touch screen according to claim 5, characterized in that: The fuzzy defect secondary tracing of the target mobile phone and the target mobile phone touch screen, and the defect solution design based on the secondary tracing result, are specifically as follows: If the actual gyroscope jitter amplitude and the actual gyroscope jitter frequency are not greater than the corresponding gyroscope standard jitter amplitude and the gyroscope standard jitter frequency, a hardware damage analysis is performed on the camera of the target mobile phone to determine whether the camera of the target mobile phone has a hardware damage problem; If so, the blur defect source of the sample image is calibrated as the hardware cause of the target mobile phone camera, and the repair plan of the damaged hardware of the target mobile phone camera is retrieved and output in the big data network; If not, then update the camera application cache and the camera application in the target mobile phone. If the sample image does not have blur defects after the camera application cache and the camera application are updated, then the source of the blur defects in the sample image is calibrated as the target mobile phone camera software cause; If the sample image still has blur defects after the camera application cache is updated and the camera application is updated, the source of the blur defect of the sample image is calibrated as a hardware problem of the target mobile phone touch screen, and the target mobile phone touch screen is replaced to ensure that the sample image does not have blur defects.

7. Image defect tracing detection system applicable to mobile phone touch screen, characterized in that: The image defect tracing detection system includes a memory and a processor, wherein the memory stores an image defect tracing detection method program. When the image defect tracing detection method program is executed by the processor, the image defect tracing detection method steps as described in any one of claims 1 to 6 are implemented.

Images