258 results about "Image noise reduction" patented technology

The invention discloses a fishing boat name identification method and system based on image processing and deep learning. The method comprises the following steps: S1, fishing boat image acquisition; s2, image preprocessing and enhancement are carried out to realize image noise reduction, deblurring and brightness adjustment processing, and image quality judgment is carried out after processing to form closed-loop optimization; s3, self-adaptive image enhancement is carried out to realize illumination compensation and image dynamic enhancement; s4, ship name area detection and segmentation; s5, ship name region feature extraction and optimization; s6, feedback and optimization of an identification result; s7, character recognition and post-processing; s8, performing correction and verification; s9, log identification and report generation; and S10, displaying and storing a result. According to the invention, the fishing boat name identification precision can be effectively improved.

The invention discloses a bearing defect detection method and system based on machine vision and ultrasonic detection, and particularly relates to the technical field of industrial automatic detection, and the method comprises the steps: S1, a synchronous collection module: carrying out pulse triggering synchronous collection, and generating a time-space reference table; s2, a feature extraction module: performing image noise reduction segmentation and ultrasonic frequency domain decomposition, and outputting a defect feature vector; s3, a fusion identification module: performing cross-modal feature alignment fusion to generate a defect classification conclusion; s4, a size measurement module: performing contour fitting to calculate inner and outer diameters, and outputting a size deviation value; and S5, a comprehensive judgment module: carrying out threshold comparison logic judgment, and generating a multi-modal detection report. According to the method, a space-time reference is established through an encoder, images are segmented in a self-adaptive mode, features are extracted through wavelet decomposition ultrasound, feature weights are re-calibrated through a parallel network and an attention mechanism, composite defects are recognized through cross-modal fusion, comprehensive judgment is conducted in combination with dimensional deviation, and a multi-dimensional quality evaluation system is achieved.

The invention discloses a medical image noise reduction and reconstruction method based on unsupervised learning, relates to the technical field of medical image processing, and solves the technical problem that an existing diffusion model is not ideal in detail recovery and noise reconstruction effect on a medical image. The method comprises the following steps: constructing at least two different first frequency domain subsets and second frequency domain subsets according to a medical image; the first frequency domain subset and the second frequency domain subset are input into a trained generation parallel network, a preliminary reconstruction image is generated, and the generation parallel network comprises at least a first generation network and a second generation network which are mutually independent; mapping the features of the preliminarily reconstructed image to a potential space, performing diffusion reconstruction in a denoising process, and encoding to obtain diffusion reconstruction features; and decoding the diffusion reconstruction features through a decoder to obtain a final reconstruction image of the medical image. According to the invention, high-frequency detail recovery of the medical image can be realized, and the high-quality medical image is reconstructed and generated.

The invention discloses a Mura defect detection method and device of a TFT-LCD display screen and electronic equipment, and relates to the field of display screens. The method comprises the following steps: acquiring an original image of a TFT-LCD display screen to be detected; preprocessing the original image to obtain a target image, the preprocessing including image noise reduction and image enhancement; based on the target image, constructing a multi-scale feature map of the Mura defect, the multi-scale feature map comprising a local mean map and a local variance map under multiple scales: performing adaptive weight fusion on the local mean map and the local variance map to obtain a confidence map of the Mura defect; according to the confidence map, utilizing morphological operation to extract a suspected defect area; performing shape analysis and contrast analysis on the suspected defect area to obtain shape features and contrast features of the suspected defect area; and screening out a target Mura defect region from the suspected defect region based on the shape feature and the contrast feature. By implementing the technical scheme provided by the invention, the accuracy of Mura defect detection is improved.

The invention discloses a photovoltaic module infrared image fault detection method based on unmanned aerial vehicle inspection, which relates to the field of fault detection, and comprises the following steps: configuring an unmanned aerial vehicle platform, planning a flight path, setting aerial photography parameters, setting an infrared thermal image acquisition and infrared image data return and storage mechanism, and performing gray normalization processing, image noise reduction, histogram equalization, edge enhancement processing and size standardization processing on the infrared image data, and performing data enhancement operation. An unmanned aerial vehicle infrared inspection technology is combined with a deep learning target detection model, a set of complete photovoltaic module infrared image fault detection process is established, and full-process automatic processing from image acquisition, image preprocessing, model detection to result evaluation and visualization can be realized. The method has the comprehensive advantages of being high in fault recognition precision, high in detection speed, standardized in processing flow and the like, and the efficiency and the intelligent level of photovoltaic power station component-level fault inspection are remarkably improved.

The invention discloses a yarn weak twist and hairiness intelligent detection system based on visual stroboscopic synchronization, particularly relates to the field of quality control in the textile industry, is used for solving the problem of accurate detection of weak twist and hairiness defects in the yarn production process, and dynamically adjusts the flash frequency of a stroboscope according to the real-time speed and acceleration data of yarn, so that the detection accuracy is improved. The light source and the yarn are ensured to move synchronously, and a high-resolution image is obtained. In combination with yarn motion spectrum analysis and image quality evaluation, a complexity score and a matching degree score are generated and are used for dynamically regulating and controlling image noise reduction and edge detection parameters, and characteristic extraction requirements under different yarn motion conditions are met. The weak twist characteristic and the hairiness texture of the yarn are accurately extracted through a multi-scale edge detection and wavelet transform algorithm, a twist distribution model is constructed, the boundary range is corrected in combination with hairiness distribution data, characteristic interference is reduced, and the detection precision and reliability are improved.

The invention relates to the technical field of artificial intelligence, is suitable for financial and medical scenes, and particularly relates to a low-illumination physical examination image noise reduction enhancement method, device and equipment and a storage medium, the low-illumination physical examination image noise reduction enhancement method comprises the following steps: obtaining a to-be-processed low-illumination physical examination image; based on a Retinex decomposition model, acquiring a preliminarily estimated illumination component of the low-illumination physical examination image, acquiring an accurately estimated illumination image according to the preliminarily estimated illumination component, and enhancing the accurately estimated illumination image through gamma transformation to obtain an enhanced image; performing noise reduction on the enhanced image based on a BM3D algorithm to obtain a noise-reduced image; and combining the enhanced image and the noise reduction image to obtain a noise reduction enhanced image. The problem that in the prior art, when enhancement processing is carried out on the low-illumination physical examination image through an image enhancement algorithm, synchronous processing is not carried out on noise and amplified noise in the enhanced physical examination image is solved, and the accuracy and efficiency of physical examination image recognition are greatly improved.

The invention relates to a system and a method for making a colored low-light image. The system for making the color low-light-level image comprises a high-sensitivity CMOS imaging module which is used for enhancing the photon collection capability and reducing the noise level to complete imaging; the image enhancement processing module adopts a histogram equalization algorithm and comprises the steps of calculating a gray histogram of the original image, calculating a cumulative distribution function according to the gray histogram and performing gray mapping to obtain an enhanced gray value; the image noise reduction processing module is used for carrying out noise reduction on the image after enhancement processing, and filtering is carried out by adopting a night image noise reduction technology, a bilateral filtering method including spatial domain noise reduction and a band-pass filter including frequency domain noise reduction; and the colorful low-light-level image generation module obtains a colorful low-light-level night vision image by adopting an interpolation algorithm and makes a colorful low-light-level image. According to the system and the method for manufacturing the colored low-light-level image, special equipment is not depended on, the imaging quality of the colored low-light-level image can be improved, and the colored low-light-level image can be manufactured.

The invention belongs to the technical field of discrete element simulation, and particularly relates to a digital twinborn discrete element model construction method based on CT (Computed Tomography) scanning, which comprises the following three steps of: in a CT scanning stage, acquiring a high-resolution cross-sectional image by optimizing scanning parameters, and carrying out image noise reduction and feature enhancement processing; in the structural analysis stage, a segmentation model is constructed based on a deep convolutional neural network, and automatic annotation of a scanned image and intelligent prediction of a three-dimensional space topological relation are realized; in the discrete element modeling stage, the reconstructed three-dimensional grid model is converted into a discrete element model with real physical attributes through parameter calibration, and discrete element simulation is achieved. According to the method, the non-destructive detection advantage of industrial CT and the feature extraction capacity of deep learning are combined, microstructure characterization is converted into macroscopic performance prediction, and a reliable simulation analysis method is provided for performance optimization of the composite material.

The invention discloses an integrated circuit scanning electron microscope image noise reduction method, which can avoid photoresist shrinkage and make image imaging clearer compared with the traditional multi-frame superposition technology in the industry. The method specifically comprises the following steps: acquiring a wafer with consistent key feature size on a semiconductor manufacturing production line; obtaining SEM images of the target point in different core particles; searching one image from the SEM images as a reference image, and taking the rest images as to-be-compared images; acquiring all overlapped regions between the reference image and the to-be-compared image; and calculating the maximum intersection area of all the overlapped areas, and carrying out pixel-level averaging on the maximum intersection area to obtain a noise-reduced SEM image. According to the method, the noise of the SEM image can be reduced, the image quality is improved, and a reliable data source is provided for product quality inspection and photoetching model driving in the semiconductor manufacturing process.

The invention relates to the technical field of image analysis, in particular to a data visualization pathological diagram analysis system which comprises a pathological image noise reduction and enhancement module, a lesion area judgment module, a local texture and global structure fusion module, a self-adaptive multi-scale segmentation module and an error correction and optimization module. According to the method, by analyzing color channels, cell structure edge features and background noise distribution in the pathological image, accurately screening noise and optimizing image filtering, the image quality is effectively enhanced, the noise is reduced, the recognition accuracy of a lesion area is improved, and the recognition accuracy of the lesion area is improved in combination with cell nucleus gradient information and tissue edge distribution. The accuracy is further improved by using gray statistics and cell density, region division is weighted and optimized through the contrast and entropy of lesion tissues, the segmentation accuracy is ensured, the image segmentation scale is accurately adjusted in combination with the cell density and color gradient information, errors are reduced, accurate segmentation of lesion regions is ensured, and the accuracy of image segmentation is improved. And finally, the precision and reliability of overall image analysis are improved.

The invention discloses an optical coherence imaging noise reduction method and device, equipment and a medium, and the method comprises the following steps: obtaining a B-scan image, and constructing a training set; training an image noise reduction model based on the deep neural network by adopting the training set; obtaining a to-be-denoised B-scan image, inputting the to-be-denoised B-scan image into the trained image denoising model for denoising, and outputting a reconstructed B-scan image; wherein the image noise reduction model comprises an encoder and a decoder, and the encoder takes a B-scan image with noise as input to obtain a potential representation of the B-scan image with noise; and the decoder takes the potential representation of the B-scan with noise as input, and outputs a reconstructed B-scan noise reduction image. According to the method, the image noise reduction model is constructed through the deep neural network, the B-scan image is subjected to noise reduction, the clear B-scan image is effectively obtained, and the method has the advantages of being excellent in noise reduction effect and high in speed. The method can be widely applied to the technical field of image imaging noise reduction.

The invention provides a radar image noise reduction processing method based on deep learning, and belongs to the technical field of machine learning, and the method comprises the steps: obtaining historical echo signal data, and determining a historical radar image data set based on the historical echo signal data; determining a training image set and a test image set based on the historical radar image data set, and constructing a noise reduction neural network model based on the training image set; evaluating the noise reduction neural network model based on the test image set, and performing second optimization on the noise reduction neural network model; and performing noise reduction processing on a real-time radar image data set determined by the real-time radar echo data based on the optimized noise reduction neural network model. The noise reduction learning ability of the model can be enhanced, the processing ability of the model to a complex noise environment can be improved, gradual fine adjustment of the model can be realized to form a performance closed loop, high-precision rapid processing of real-time radar data can be realized, the denoising effect of a radar image can be improved, and the noise reduction precision in the complex environment can be improved.

The invention discloses a fluorescence image adaptive enhancement noise reduction method and a fluorescence endoscope imaging system, and relates to the image processing related field, and the method comprises the steps: loading a fluorescence endoscope image; dividing the fluorescent endoscope image to obtain D window fluorescent images; extracting a fluorescence image of the d-th window; filtering and noise reduction are carried out to generate a d-th window noise reduction image; performing noise reduction verification to generate a d-th window noise reduction verification image; continuously performing filtering noise reduction and noise reduction verification on each window fluorescence image in the D window fluorescence images to generate D window noise reduction verification images; and carrying out image fusion on the D window noise reduction verification images to generate a fluorescence endoscope noise reduction image. The technical problem that the quality of the fluorescence endoscope noise reduction image is uneven due to the fact that an existing fluorescence endoscope image noise reduction method is not ideal in noise reduction effect and lacks a unified verification process is solved, and the technical effect of improving the quality of the fluorescence endoscope noise reduction image is achieved.

The invention relates to the technical field of image noise reduction and enhancement, and discloses a CT image noise reduction and enhancement system and method which are used for providing a data interface module of a data transmission channel between the system and imaging equipment. The original data preprocessing module is used for preliminarily correcting the collected original projection data; the noise simulation module is used for establishing a noise model to simulate noise distribution in actual work according to physical characteristics and an imaging principle of the X-ray detector; the noise suppression simulation module is used for reducing noise based on a noise suppression algorithm of a physical model; the deep learning fusion module is used for optimizing the system in combination with a physical model noise reduction method and a deep learning technology; the image protection module is used for particularly paying attention to protection of important anatomical structures in the noise reduction process and avoiding loss of key diagnosis information due to excessive noise reduction; and the post-processing enhancement module is used for improving the image quality after the image processing is completed and ensuring the clinical readability of the image.

The invention provides a shale pore type identification method based on a digital core, and the method comprises the steps: carrying out the gray adjustment processing and image noise reduction processing of an original gray image of a to-be-identified shale sample, and obtaining a standard gray image; organic matter and transition zone images in the standard grayscale image are extracted, organic matter identification processing is carried out, and an organic matter connected body image is determined; extracting a pore image in the standard gray-scale image, performing pore recognition processing in combination with the organic matter connected body image, and determining an organic pore image and an inorganic pore image; and performing micro-crack identification processing on the inorganic pore image, and determining a micro-crack image and an inorganic pore image. By applying the method, the organic pores, the inorganic pores and the microcracks of the shale can be quantitatively identified, the pore structure of the shale can be quantitatively represented, and the structural parameters of the organic pores, the inorganic pores and the microcracks can be analyzed.

The invention provides a gamma imaging noise reduction detector based on depth information, which comprises a layered probe, a host and an acquisition circuit, and is characterized in that the layered probe is of a multi-layer structure formed by arranging single-layer position sensitive probes front and back; the acquisition circuit acquires timestamps, positions and amplitude information of the electric pulse signals; the host is responsible for collecting the number of the collection circuit and generating a detector image, forming the same particle event by using the electric pulse signals with the same timestamp, and determining thickness direction information of activated pixel data corresponding to each electric pulse signal through the number of the collection circuit. And only the data of the foremost activation pixel in the thickness direction is accumulated into the detector image. According to the method, a thick detector is divided into layered probes to obtain depth information of particle energy deposition, so that the noise influence caused by scattering in a detector crystal in medium-high energy gamma ray imaging is reduced, a first scattering point is speculated to eliminate secondary scattering noise of medium-high energy gamma rays, and the contradiction between the imaging speed and the imaging quality is solved.

The invention discloses a cement homogeneity detection method based on image processing. The method comprises the following steps: S1, image preprocessing: converting an original image into a grayscale image; s2, image noise reduction: salt and pepper noise is eliminated through a filtering technology, and the influence of the salt and pepper noise can be effectively eliminated; s3, image enhancement: carrying out sharpening and detail enhancement on the grayscale image by adopting a Laplace enhancement algorithm; s4, carrying out threshold segmentation: carrying out threshold segmentation through gray difference, and identifying an interested part which is not uniformly stirred; s5, watershed segmentation of the adhesion part: identifying a plurality of local maximum gray value points by adopting a watershed segmentation method based on extended maximum transformation, and combining the local maximum gray value points into a unique maximum value point to realize accurate segmentation of the adhesion area; and S6, homogeneity judgment. According to the invention, through an image processing algorithm, the number, size, distribution and other characteristics of the granular cement blocks are accurately analyzed, so that whether the cement is uniformly stirred is judged.

The invention discloses a neural network structure search method for self-supervising image noise reduction, and belongs to the field of neural network structure search and image noise reduction. According to the method, a search space of a multi-scale encoder is designed, so that a network structure can fuse feature information extracted from different scales, more image details can be recovered, and the network structure with a better noise reduction effect can be found; based on a search space of a multi-scale encoder, an improved genetic algorithm is provided to search a network structure, a coarse-grained population initialization method, a distance-guided parent selection method and a modularization-based crossover mutation operation are designed, and the search capability of the genetic algorithm is improved. Compared with an existing self-supervised image noise reduction method and a neural network structure method, a better noise reduction effect can be achieved; experimental results prove that the self-supervised image noise reduction neural network structure searching method can effectively discover a network structure with a better noise reduction effect.

The invention provides a training method for an image noise reduction model, and relates to the technical field of image processing, and the method comprises the steps: obtaining a plurality of first-class sample pairs in a target training set; inputting the noise image in each first sample pair and the corresponding image quality score into a to-be-trained target image noise reduction model to obtain a denoising result corresponding to the noise image in each sample pair; determining a target loss value of the target image noise reduction model based on the denoising result corresponding to the noise image in each first type of sample pair and the target image in each first type of sample pair; and in response to determining that the target image noise reduction model is not converged according to the target loss value, performing parameter adjustment on the target image noise reduction model, and returning to the step of obtaining the plurality of first-class sample pairs in the target training set. According to the scheme, the noise reduction effect of the image noise reduction model obtained through training in an actual scene can be improved.

The invention provides an unsupervised low-dose photon counting CT (Computed Tomography) reconstruction method and an unsupervised low-dose photon counting CT reconstruction device based on optimal transmission. According to the method, a low-dose photon counting CT reconstruction problem is modeled as a Cantrovitch problem under a distribution consistency constraint, and the method comprises the following steps: constructing and training an optimal transmission-based unsupervised projection recovery network and constructing and training an optimal transmission-based unsupervised image noise reduction network; performing detail recovery on the photon counting CT low-dose projection by using the trained unsupervised projection recovery network to generate a denoised projection; reconstructing the denoised projection by using an image reconstruction algorithm to generate an intermediate image; and performing noise reduction on the intermediate image by using the trained unsupervised image noise reduction network to obtain a final de-noised image. According to the method, the problem that a high-quality pairing data set is difficult to obtain and a photon counting CT low-dose image is difficult to reconstruct can be solved.

The invention relates to the technical field of image recognition anti-counterfeiting, in particular to an authenticity identification method for analyzing micro-stereoscopic structure distribution on the surface of a product by using a mobile phone deep learning model. Comprising the following steps: image acquisition; image preprocessing: performing image noise reduction and region segmentation operation on the acquired image, extracting the edge of the micro-stereo structure through a Canny edge detection algorithm, performing fitting to generate a closed contour, and positioning an effective detection region of the micro-stereo structure in combination with contrast enhancement processing; extracting features; performing feature comparison; and outputting a result. According to the method, the limitation of the traditional one-dimensional or two-dimensional anti-counterfeiting feature is broken through by extracting the multi-dimensional features such as the height difference, the spatial arrangement and the geometrical shape of the micro three-dimensional structure on the surface of the product. The features are uniquely combined by depending on the three-dimensional distribution characteristic of the micro-stereo structure, and the complexity of the features enables high-precision equipment to be difficult to accurately imitate and copy, so that the anti-counterfeiting reliability is improved.

The invention discloses an image adaptive enhancement method and system based on deep learning, and relates to the field of remote control. The stamping plate image acquisition module is used for acquiring an image of a to-be-detected stamping plate of a production line through an industrial camera; the deep learning-based image adaptive enhancement method and system comprises the following steps: collecting an image of a stamped plate: storing and saving the image of the stamped plate through an industrial camera by the system; according to the step of image processing, due to the fact that noise and uneven illumination exist in collected stamping plate images, image noise reduction and filtering algorithm processing are carried out on the images, and stamping plate surface defect detection is mainly carried out on the input stamping plate images through a defect classification algorithm. And information about whether the surface of the stamped plate has defects and defect positions is stored.

The invention provides a method and a system for removing calcification artifacts of a CT (Computed Tomography) image. The method comprises the following steps of: preprocessing the CT image; semantic feature information extraction is carried out on the preprocessed CT image; fusing the preprocessed CT image and the semantic feature information, and inputting the fused image and semantic feature information into a GAN-based artifact removal model to eliminate calcification artifacts so as to realize noise reduction; an improved perception extraction module is introduced into a generator, and a global / local-semantic judgment module is added into a discriminator, so that the image noise reduction effect after the calcification artifacts are removed is further improved. By means of the GAN-based artifact removal model, common noise reduction can be effectively achieved, meanwhile, calcification artifacts in the CT image are removed, so that the quality of the CT image is improved, the judgment accuracy of a doctor for the blood vessel stenosis degree is further improved, and then the misdiagnosis rate of CT diseases is reduced.

The present application provides an image denoising method, device, equipment and storage medium, belonging to the field of image processing technology, the method comprising: determining a first image to be subjected to denoising; determining the noise level of the first image and the gradient value of each pixel in the first image; determining a gradient threshold according to the noise level of the first image; determining a diffusion function corresponding to each pixel in the first image according to the gradient value of each pixel in the first image and the gradient threshold; performing denoising on the first image according to the diffusion function corresponding to each pixel in the first image to obtain a target image. The step effect in the target image is avoided, and the accuracy of the denoising is improved.

The invention provides an image noise reduction enhancement method based on an additive Gaussian diffusion model, which comprises the following steps: selecting a plurality of clear images, gradually adding additive Gaussian white noise to the clear images until the images are completely polluted by the noise, and generating a series of noise images with different pollution measures; forming a training data set by the noise images with different pollution measures, the corresponding time steps and the global noise samples obeying standard normal distribution, and training a U-Net network; establishing an additive Gaussian diffusion model for reversely deriving and estimating a noise image at the previous moment; collecting a noise image to be denoised, obtaining an initial value of the noise level of the noise image, and calculating a time step number corresponding to the initial value of the noise level through a time coding function; and inputting a polluted image to be subjected to noise reduction and the corresponding time step number into the trained U-Net network, connecting a vertical additive Gaussian diffusion model in parallel, and removing the noise of the noise image by adopting a loop iteration optimization strategy through back diffusion calculation.

The invention discloses an electrical equipment test data acquisition method and system based on image recognition, and relates to the field of electrical equipment test data acquisition methods, and the method comprises the steps: obtaining an electrical test equipment nameplate instrument image; image deformity correction; image region segmentation; performing image noise reduction; image enhancement and character image segmentation; training a character recognition model; checking and correcting the characters; a test data automatic input system; the electrical equipment test data acquisition system based on the image recognition method helps electrical equipment testers to quickly recognize complex electrical test data and detect field test device data through a mobile terminal APP, replaces a traditional manual writing and recording mode, and solves the problems that field data is inconvenient to transcribe, and the test efficiency is high. Data mistakes and omissions caused by nonstandard writing, writing errors and the like of personnel are avoided from the source, and the accuracy of test technical data is ensured; compared with an original mode, labor cost is reduced, and working efficiency is improved.

The invention relates to an image noise reduction technology, and discloses an image noise reduction method based on distribution standardization alignment, which comprises the following steps: calculating the total number of pixels of an image to be processed, carrying out linear transformation on pre-acquired standard normal distribution and the total number of pixels to obtain normal gray level distribution, calculating the initial gray level distribution of the image to be processed, and constructing a lookup table and a mapping table in each pixel value domain according to the initial gray level distribution and the normal gray level distribution, performing standardized adjustment on pixel values of the to-be-processed image by using the lookup table to obtain a standardized image, and performing noise reduction processing on the standardized image by using a pre-trained noise reduction model to obtain a noise reduction result of the to-be-processed image. And carrying out gray level distribution inverse mapping on the noise reduction image to obtain a final output image. The invention further provides an image noise reduction system based on distribution standardization alignment, electronic equipment and a storage medium. According to the invention, the image noise reduction effect can be improved.

The invention relates to a low-dose SPECT image noise reduction method and system based on an unsupervised strategy and a generalized diffusion model. The method comprises the following steps: carrying out slicing processing on a low-dose SPECT image to obtain a training image; in a forward propagation stage of the unsupervised diffusion training model, determining an attenuation coefficient degradation operator according to the body mass index of the patient, and carrying out image noise addition to obtain a random Poisson noise image; in a back propagation stage of the unsupervised diffusion training model, extracting optimization factors and time step length features, and performing denoising processing on the random Poisson noise image to obtain a predicted low-dose SPECT image; and loading the optimal weight into a back diffusion inference model, carrying out back diffusion by taking the predicted low-dose SPECT image as a diffusion end point, and carrying out noise reduction processing on the predicted low-dose SPECT image to obtain a predicted full-dose SPECT image. The data preparation cost can be reduced, artifacts are reduced, and the denoising effect is improved.

The invention relates to the technical field of metal detection, and discloses a polarization imaging metal surface scratch identification system. The system comprises a polarization imaging unit, an image preprocessing unit, a polarization feature extraction unit, a scratch recognition unit and a parameter self-adaption unit. The polarization imaging unit collects and transmits a metal surface polarization image; the image preprocessing unit performs noise reduction and enhancement on the image to generate an optimized image; a polarization feature extraction unit analyzes polarization state parameters of the optimized image to obtain polarization degree and azimuth angle distribution, and extracts polarization features of the scratch candidate area; the scratch identification unit compares the polarization characteristics with a preset template to generate an identification result; and the parameter adaptive unit dynamically adjusts the acquisition parameters of the polarization imaging unit according to the definition and contrast ratio of the scratch in the recognition result. According to the system, the polarization imaging technology is utilized, multi-unit cooperative work is combined, metal surface scratches can be accurately recognized in a complex environment, the system adapts to different detection conditions, and the recognition effect is improved.