299 results about "Defective pixel" patented technology

The invention relates to the technical field of defect detection, in particular to an efficient identification and detection method for printing defects on the surface of a metal plate. The method comprises the following steps: acquiring a difference characterization value and a suspected judgment index value of a pixel point to be detected, and acquiring a suspected defect pixel point according to the suspected judgment index value; clustering all the suspected defect pixel points according to the suspected judgment index values and the coordinate values of the suspected defect pixel points to obtain each cluster; according to the difference characterization value of the suspected defect pixel point, the gradient direction of the suspected defect pixel point in the cluster to which the suspected defect pixel point belongs and the gray value standard deviation and mean value of all the suspected defect pixel points in the cluster to which the suspected defect pixel point belongs, obtaining an abnormal degree characterization value of the suspected defect pixel point; and performing printing defect identification on the surface of the to-be-detected printed metal plate according to the abnormal degree characterization value. And the identification accuracy of the printing defect area on the surface of the printing metal plate can be improved.

The invention discloses a display screen dead pixel fault prediction detection method and system. The method comprises the following steps: collecting a visible light image, a short wave infrared image, an ultraviolet fluorescence image, a driving current time sequence signal and surface temperature distribution data of a display screen; performing dynamic noise injection and cross-modal alignment on the data to generate a preprocessed multi-modal data set; inputting the preprocessed multi-modal data set into a space-time fusion network, extracting space-time correlation characteristics in a dead pixel diffusion process through the space-time fusion network, and generating a pixel-level dead pixel mask of a current frame and a diffusion probability thermodynamic diagram of a future time period based on the space-time correlation characteristics; and according to the dead pixel mask and the diffusion probability thermodynamic diagram, dynamically adjusting detection parameters to adapt to target equipment, and outputting a probabilistic diagnosis result of dead pixel causes in combination with physical relevance of multi-modal data.

The embodiment of the invention provides a defect detection method, electronic equipment, a storage medium and a program product, and the method comprises the steps: obtaining the actual morphology data of one or more local regions of a target wafer and a global layout of the target wafer, and carrying out the global layout of the target wafer for each local region based on the mapping relation between the wafer morphology and technological parameters, determining a prediction process parameter corresponding to the actual morphology data of the local region, and predicting the wafer morphology of the global layout under the prediction process parameter of the local region for each local region based on a mapping relationship between the wafer morphology and the process parameter to obtain wafer morphology prediction data, and determining the defective pixel areas of the target wafer based on the predicted wafer morphology prediction data corresponding to each local area, determining a to-be-detected area from all the defective pixel areas, and scanning the to-be-detected area of the target wafer by using defect scanning equipment to obtain a defect detection result, so as to effectively improve the efficiency and precision of defect detection.

The embodiment of the invention discloses a brightness compensation method and device for a display panel and a display device, and the method comprises the steps: determining a to-be-compensated pixel corresponding to a defective pixel according to the type of the defective pixel of the display panel, and enabling the to-be-compensated pixel to be corresponding to the defective pixel according to the type of the defective pixel during the subsequent compensation of the display panel, and performing brightness compensation on the to-be-compensated pixel corresponding to the defective pixel so as to realize brightness compensation on the defective pixel. The types or compensation modes of the to-be-compensated pixels corresponding to the at least two different types of defective pixels can be different, so that targeted brightness compensation is performed on the defective pixels in the display panel, the display effect of the display panel is improved, and the user experience is improved.

The invention provides an intelligent POB module integration platform and a control method thereof, and the method comprises the steps: detecting the state data of each pixel point of an LED display screen through an optical sensing array, obtaining a damaged pixel point in the LED display screen, recording the damaged pixel point to a storage unit, and maintaining a dead pixel position table in the storage unit; compensating the brightness of the bad pixel points through the neighborhood pixel points; the display content of the LED display screen is predicted through AI during compensation, and the refresh rate of the LED display screen is dynamically driven according to the predicted display content; according to the physical pixel density, the LED display screen is divided into grid partitions, the grid partitions are sorted according to priorities, the real-time power consumption of each partition is calculated after the LED display screen is refreshed, and if the real-time power consumption is larger than a preset threshold value, the current and the duty ratio of the low-priority area are reduced. Through the control method, a compensation scheme can be started in time after a dead pixel occurs, the maintenance and replacement cost is reduced, and the operation and maintenance efficiency is improved.

The invention provides a Mini led display screen dead pixel detection and positioning method, device and system based on deep learning, and the method comprises the steps: applying a preset driving signal sequence to a pixel array of a Mini led display screen line by line, obtaining the brightness feedback data of each pixel in different driving stages, obtaining a brightness response sequence, and carrying out the detection and positioning of the dead pixel of the Mini led display screen. Carrying out spatial recombination processing on the brightness response sequence according to a physical arrangement structure of the pixel array to generate a brightness response matrix, carrying out local region feature difference analysis on the brightness response matrix, identifying a brightness response abnormal region, generating a candidate dead pixel position set, and obtaining a candidate dead pixel position set; and calling a pre-trained dead pixel recognition network to carry out feature learning processing on the candidate dead pixel position set, outputting a dead pixel probability corresponding to each candidate dead pixel position, screening the candidate dead pixel position set based on the candidate dead pixel probability, removing candidate positions with the dead pixel probability lower than a preset condition, and generating a dead pixel positioning result set. According to the invention, the accuracy and reliability of detecting and positioning the defective pixels of the Mini LED display screen can be effectively improved.

The invention discloses a display device and a defect pixel compensation method of the display device, the display device comprises a display panel and a processor, the processor determines the defect type of a defect sub-pixel and an initial gray scale value corresponding to the defect sub-pixel; determining at least one target gray-scale value according to the defect type of the defective sub-pixel, the initial gray-scale value and a plurality of to-be-displayed gray-scale values corresponding to the defective pixel to which the defective sub-pixel belongs; based on the at least one target gray-scale value, updating the to-be-displayed gray-scale values corresponding to other sub-pixels in the defective pixels; and the display panel displays the to-be-displayed image after the plurality of to-be-displayed gray-scale values corresponding to the defect pixels are updated. According to the target gray-scale value, the to-be-displayed gray-scale values of other sub-pixels except the defective sub-pixels in the defective pixels are updated, so that the display color cast problem caused by the defective sub-pixels is compensated, the perceptibility of the defective sub-pixels in human eyes is reduced, and the display effect of the display device is improved.

The invention discloses a method and system for detecting surface defects of a sheet metal part before coating, and the method comprises the following steps: S1, multi-modal data collection: synchronously collecting three-dimensional point cloud data and a hyperspectral image of the surface of the sheet metal part through a laser three-dimensional contourgraph and a hyperspectral camera, and recording environment illumination and temperature parameters at the same time; s2, data preprocessing: performing denoising and smoothing processing on the three-dimensional point cloud, extracting surface contour features, and performing defective pixel repair, waveband selection and reflectivity correction on the hyperspectral image; according to the surface defect detection method and system before coating of the sheet metal part, through multi-modal data acquisition and fusion of the three-dimensional morphology and the hyperspectral features, compared with traditional single visual detection, the surface defects of the sheet metal part can be more comprehensively and accurately recognized, particularly, tiny defects and oxide layer defects which are difficult to distinguish by naked eyes, the recognition rate is remarkably increased, and the detection accuracy is improved. And by adopting the lightweight deep learning model, the detection precision is ensured, and the reasoning speed is improved at the same time.

The invention discloses a dead pixel detection method and system combining spatial domain analysis and time domain analysis, and belongs to the technical field of image processing. The method comprises the following steps: an airspace analysis stage: extracting edge information through a Sobel operator; carrying out image binaryzation, filling potential holes by using morphological closed operation, and strengthening dead pixel areas; identifying all candidate dead pixel regions by adopting a connected domain labeling algorithm; in the time domain analysis stage, the spatial position and area characteristics of candidate dead pixel areas are combined, the candidate dead pixel area corresponding relation between cross frames is established, the continuity of the same candidate dead pixel area on a time axis is judged, and only two assumptions are met, the time domain analysis is completed. And finally judging the candidate dead pixel regions with the areas and the contrast ratios of the dead pixel regions exceeding the set threshold values as stable dead pixels and outputting the stable dead pixels. By analyzing the stability of the dead pixels in the time sequence, abnormal dead pixels continuously existing in multiple frames can be effectively recognized, and particularly, the method is more stable and accurate when detecting small-area and low-contrast dead pixels.

The embodiment of the invention provides a display screen direct-drive display method based on a data algorithm, and belongs to the field of display screen display methods. The method comprises the following steps: acquiring RGB frame data and environment sensing data; preprocessing the acquired RGB frame data, generating an importance degree matrix, and grading the image area of the display screen; dynamically distributing the number of sub-cycles of pulse width modulation, matching different pulse widths for image areas of the graded display screen, and generating a compensated gray matrix by using a nonlinear compensation method; generating a dead pixel position matrix according to the scanned pixel state of the display screen, and generating a compensated brightness matrix of the display screen by using a brightness diffusion compensation method; a plurality of logic boards are dynamically synchronized by utilizing an FPGA, and a dead pixel avoidance mechanism is adopted to realize direct-drive display of a display screen. According to the invention, pocking marks are eliminated, low gray scale maintains high refresh rate, chromatic aberration is reduced, dead pixels of the display screen do not need to be halted for maintenance, cost is reduced, brightness is compensated for natural transition, and visual uniformity is compensated for.

The invention relates to the field of image processing, in particular to a thermosensitive coating defect detection method and system for thermosensitive paper, and the method comprises the steps: collecting a historical thermosensitive coating original image with defect marks; the defects are clustered, the size of the center point of any cluster serves as the size of an initial window, grid division is conducted on a thermosensitive coating grey-scale map obtained in real time, and a gray-scale co-occurrence matrix of each grid is obtained; taking any grid as a central grid, calculating the sum of differences between the central grid and the gray-level co-occurrence matrixes of all other grids in a preset neighborhood, and normalizing the sum of differences to obtain a texture difference matrix of the central grid; and calculating the probability that the grid has texture defects, calculating the integrated defect probability of pixel points in the grey-scale map, and if the integrated defect probability of the pixel points is greater than a preset threshold value, judging that the pixel points are defective pixel points. According to the invention, potential defects can be effectively detected from the thermosensitive coating image, so that the accuracy of defect detection is improved.

In the embodiment, the location of the defective pixel can be determined by measuring the operating current of each LED of the pixel, and the sub-LED can be driven in response to the location information of the defective pixel, thereby implementing normal LED operation even in a pixel defect situation.

The invention relates to the technical field of dead pixel detection and compensation of a digital light processing projection system, in particular to a DLP dead pixel calibration and compensation method and system based on a Transform attention mechanism. According to the DLP dead pixel calibration and compensation method based on the Transform attention mechanism, a multi-scale feature pyramid is constructed, global features are extracted by using a layered multi-mode Transform encoder, and a dead pixel confidence coefficient heat map is generated in combination with a dynamic sparse mask multi-head attention mechanism; dynamically selecting compensation strategies for different types of dead pixels by adopting a dead pixel type-scene demand dual-drive time-space domain compensation pipeline; ambient light and partition temperature sensor data are integrated, and the influence of optical crosstalk and thermal drift is eliminated through a brightness adaptive calibration function; and finally, carrying out closed-loop verification and iterative optimization by adopting a loss function fused with HVS perception loss. According to the method, the defective pixel detection precision is improved from 82.3% to 99.2%, the PSNR after compensation is improved from 32.6 dB to 42.5 dB, the processing delay is reduced to 7ms / frame, and the visual quality and reliability of a DLP system are improved.

The invention provides a multi-camera curved screen Demura compensation data dynamic splicing method and system, and relates to the technical field of display, and the method comprises the steps: carrying out the flat field correction, dark current correction and dead pixel correction of a plurality of cameras, and obtaining the cameras after uniformity correction; generating a positioning map, collecting the positioning map through a plurality of corrected cameras, and calculating a sub-pixel coordinate mapping relation corresponding to each camera; dynamically estimating the curvature change of the curved screen, constructing a curved surface calibration model and generating a curved surface correction coefficient; performing brightness range correction on the curved surface correction coefficient to obtain a corrected dynamic curved surface correction coefficient; and brightness correction is carried out on gray scale images acquired by the plurality of cameras, and seamless data splicing is realized through brightness calibration among the cameras. The method effectively solves the problem of data acquisition and splicing of the large-size curved screen, is suitable for various display devices, reduces the hardware cost and the process complexity, and has high precision, flexibility and wide application prospects.

The invention relates to the technical field of infrared detection, and discloses an infrared detector dead pixel detection method, which comprises the following steps: acquiring a current scene image during the working period of an infrared detector and a plurality of frames of isothermal uniform surface background images acquired by the infrared detector at different working temperatures to form a background image sequence; performing singular point detection on each frame of image in the current scene image and the background image sequence by adopting an improved wavelet transform algorithm, screening out effective singular points, and calculating a singularity value of each effective singular point; based on the singular point detection result, the singularity value difference and historical background image data, static dead pixels, dynamic dead pixels and noisy points are distinguished through a multi-layer judgment rule, and a dead pixel detection result is obtained; and periodically evaluating the accuracy of the defective pixel detection result, and dynamically calibrating the detection parameters according to the evaluation result. Through collaborative analysis of the current working image and the multi-temperature background image sequence, dead pixels can be accurately detected in real time in the normal working process of the detector.

The invention provides a panel defect detection method and system based on multi-model fusion, and relates to the technical field of panel defect detection, and the method comprises the steps: carrying out the first-stage defect detection of a panel sample image through employing a mode of combining edge detection and isolated forest determination, so as to obtain the defect of a screen display abnormality type; performing secondary defect detection on the panel sample image by adopting an image classification model to obtain defects of a screen scratch category; performing three-stage defect detection on the panel sample image by adopting the target detection model to obtain defects of screen dead pixels and bad line categories; and performing result fusion on different types of defects to obtain a panel defect detection result. According to the invention, a mode of combining an image processing technology, a machine learning technology and a neural network technology is adopted, and the problems that the defect of panel badness cannot be fully detected and missing detection is easy to occur due to the fact that an existing single algorithm or model based on a machine vision technology is difficult to cover multiple types of bad mobile phone panels are solved.

The present application discloses an image reconstruction method, an apparatus, a device, a medium, and a product. According to the image reconstruction method in the embodiments of the present application, all Swin Transformer modules of an encoding module are connected by means of a patch merging layer; all Swin Transformer modules of a decoding module are connected by means of a patch expanding layer; the encoding module and the decoding module are symmetrical in structure; and finally, an upsampled feature map is input into a reconstruction module so as to obtain an RGB image by means of conversion. By constructing a network on the basis of a U-Net structure, the present solution is able to restore a corresponding high-resolution RGB image from a feature map, and interpolate missing values and defective pixels in a RAW image captured by an event camera while performing demosaicing, thereby obtaining a high-quality RGB image. Moreover, the use of Swin Transformer architecture as a feature extractor allows for the effective capturing of long-range dependencies, and ensures that consistency of local and global information is maintained.

The invention relates to the field of color sorter control, in particular to a control algorithm and system for adjusting air blowing of a spray valve based on shapes and dead pixel positions, and discloses the control algorithm for adjusting air blowing of the spray valve based on shapes and dead pixel positions, which comprises the following steps of: performing image analysis on a falling object through an image analysis module; obtaining image information of the object; the image information is recognized through an image recognition module, shape features and dead pixel positions of the object are obtained, and according to the shape features and the dead pixel positions, the to-be-removed position of the object and the offset distance of the mass center point relative to the center point of the external rectangle are obtained; and the blowing position of the spray valve is determined according to the to-be-removed position and the offset distance, blowing parameters are adjusted according to the blowing position of the spray valve, and removing operation is executed. Through cooperation of the image analysis module and the image recognition module, the blowing parameters of the spray valve can be dynamically adjusted, and it is guaranteed that the to-be-removed object can be accurately removed.

A display device includes: a display panel including a plurality of pixels; a sensing circuit to generate sensing values of the plurality of pixels by sensing characteristics of the plurality of pixels; and a panel defect normalizer to compensate for the sensing values, the panel defect normalizer including: a preliminary defective pixel detector to detect a preliminary defective pixel; a defective pixel detector to detect a defective pixel; a defective line detector to detect a defective line; a defective line compensator to compensate for a sensing value of the defective line; and a defective pixel compensator to compensate for a sensing value of the defective pixel.

The invention relates to the technical field of display screen detection, and discloses a display screen dead pixel positioning method based on image feature matching, and the method comprises the following steps: collecting image data of a display screen in different test modes to generate a feature matrix, and constructing a dead pixel detection template set; setting a reference parameter type, obtaining a pixel distribution characteristic, a brightness gradient and a chrominance offset of a historical image, and establishing a final matching degree associated equation; extracting a pixel distribution characteristic and a brightness gradient parameter of a current display screen, and generating a characteristic matching set and a brightness parameter set by combining equation optimization; and updating detection logic to generate a target positioning scheme, calibrating a pixel coordinate mapping relation, and carrying out binarization processing on calibrated coordinates to generate a dead pixel position marking matrix. According to the method, through multi-mode feature analysis, parameter dynamic association and coordinate calibration, the efficiency and accuracy of dead pixel positioning are improved, and the method is suitable for dead pixel detection of various display screens.

The invention discloses an outdoor dot matrix display screen picture flaw detection method, device and system based on texture features and a server. According to the method, at least two pure color test pictures are obtained, and defect full-coverage detection is realized by utilizing sensitivity differences of different color channels; besides, multi-scale feature extraction is carried out based on a Gabor filter, so that scale features such as single-point dead pixels, regional dark spots and module-level distortion can be captured, and multiple types of flaws such as brightness deviation and large-area uneven brightness can be covered. According to the method, the detection rate of traditional leak detection scenes such as small dead pixels and color deviation is increased to 95% or above, the environment robustness is high, the false detection rate in the complex illumination and vibration environment is low, and a high-reliability automatic defect detection solution is provided for the field of outdoor advertisements and public display.

The invention discloses a pixel response inconsistency correction method and system. The method relates to the technical field of pixel response correction and comprises the following steps of pixel dynamic offset calculation, flat field coefficient acquisition, pixel-level comprehensive correction and dead pixel self-repairing secondary optimization. The method comprises the following steps: acquiring a dark frame image and real-time temperature data, and calculating a dynamic offset of a pixel by means of a temperature-time double-domain prediction model; a flat field coefficient is obtained by fusing a basic value calibrated by a laboratory miniature integrating sphere in a full spectrum band and an updated value obtained by in-orbit observation of a deep space and a cloud top and processing of dead pixel detection and the like; then combining a GPS, an attitude angle and a real-time temperature, and using an offset and a flat-field coefficient to carry out pixel-level correction; and finally, a gain compensation matrix is output by means of a deep reinforcement learning model containing a convolutional neural network and a long-short-term memory network, the image is secondarily optimized through a dead pixel self-repairing module, the pixel response inconsistency correction precision is improved, and the problem that in the prior art, the pixel response inconsistency correction error is large is solved.

The invention provides a method for correcting photoetching simulation dead pixels. The correction method comprises the following steps: providing a first mask layout, and obtaining simulation dead pixels in the first mask layout and positions of dead pixel areas where the simulation dead pixels are located; obtaining a design layout corresponding to the dead pixel area as a target graph, and correcting the boundary of the target graph to obtain a virtual graph; performing optical proximity correction on the virtual graph to obtain a corrected graph; and performing photoetching analogue simulation on the corrected pattern, and judging whether a simulation dead pixel exists or not based on an evaluation function, wherein parameters of the evaluation function comprise an edge placement error and key size change. In this way, the convergence speed of the simulation dead pixels can be increased, the restoration accuracy of the simulation dead pixels can be guaranteed, and the overall efficiency of optical proximity correction is improved.

Provided is a photodetection device capable of suppressing generation of defective pixels. Specifically, the photodetection device includes a semiconductor substrate on which a plurality of photoelectric conversion units is formed; and a plurality of optical filters disposed on a light incident surface side of the semiconductor substrate. Furthermore, each of the optical filters includes a metal structure including a metal material of the same kind. Moreover, the photodetection device includes, between the metal structures, a slit portion spatially sectioning the metal structures.

The invention provides a wafer-level inertial device test path planning method and system, and relates to the technical field of inertial device detection.The method comprises the steps that whether a to-be-tested point position capable of being tested exists or not is judged by determining a test point position coordinate of an index path and the type of a multi-view probe; according to the type of the multi-view probe, judging whether defective pixels exist in a plurality of point position coordinates of a neighborhood of the point position to be detected or not; moving the to-be-detected point position, judging whether a plurality of point position coordinates in the neighborhood of the moved to-be-detected point position have dead pixels or not, skipping the moved to-be-detected point position if the dead pixels exist, and re-determining the to-be-detected point position of the index path; and if the dead pixel does not exist, sending a moving instruction to the probe station, connecting the multi-view probe with the point location to be tested and a plurality of point locations in the neighborhood, and performing driving end testing and detection end testing. According to the method, in the batch test process of the wafer-level inertial device, path planning can be carried out, dead pixels are avoided in real time, the test efficiency is improved, and the test cost is reduced.

The invention provides a mask pattern optimization method, a medium, a product and equipment. The method comprises the following steps: acquiring a mask pattern to be corrected and a corresponding ideal photoetching pattern; calling an artificial intelligence model in the inversion photoetching tool to generate a photoetching simulation result according to the mask pattern to be corrected; calculating a cost value according to the photoetching simulation result and the ideal photoetching graph; correcting the mask pattern to be corrected based on the cost value by using an inversion photoetching algorithm to obtain a corrected mask pattern; and taking the corrected mask pattern as a to-be-corrected mask pattern, and repeatedly executing the step of calling the artificial intelligence model in the inversion photoetching tool to generate a photoetching simulation result according to the to-be-corrected mask pattern until a preset cycle ending condition is met, thereby obtaining an optimized mask pattern. Through the method, the inversion photoetching tool can be highly matched with an OPC tool of a corresponding manufacturer, so that a process window of a dead pixel area is effectively improved, and a solid foundation is laid for improvement of the overall yield of technical nodes.

The invention relates to a camera pixel dead pixel detection and correction method and system, a terminal and a medium, and relates to the technical field of image detection and correction, and the method comprises the steps: determining a plurality of light intensity points based on the maximum light intensity responded by a camera pixel, and controlling a uniform light source to irradiate a camera to obtain an image data set; calculating a plurality of response coefficients corresponding to each pixel based on the image data set; performing dead pixel discrimination on each pixel of the camera based on the first response coefficient matrix, the second response coefficient matrix and the goodness-of-fit matrix to generate a dead pixel matrix; and for each pixel, determining the type of the pixel based on a dead pixel matrix, and correcting the gray value of the pixel based on the type and position of each pixel and a preset dead pixel correction algorithm to obtain a corrected image data set. According to the invention, the dead pixel of the camera pixel can be accurately detected, the pixel gray value is effectively corrected, and the image quality is improved.

The invention relates to the technical field of liquid crystal screen detection with a defective pixel classification function, and discloses a liquid crystal screen detection method and device with the defective pixel classification function, and the device comprises a classification base, a conveying belt arranged at the top of the classification base, a second conveying belt arranged at the top of the classification base, and a third conveying belt arranged at the top of the classification base. The surface of the classification base is fixedly connected with a support, a cleaning and detecting device is arranged above the classification base, a classification device is arranged above the classification base, and a collecting device is arranged on the surface of the classification base. When the conveying belt is started, the liquid crystal display above the conveying belt can be driven to move, when the liquid crystal display passes through the position below the cleaning soft brush, the cleaning soft brush can clean dust above the liquid crystal display, and it is avoided that the accuracy of machine detection is affected by the dust; and dust cleaned by the cleaning soft brush can enter a dust collection box through a dust collection cover through a dust collection pipe, and collection work is carried out.

This application relates to image processing, and provides a defect detection method, apparatus, electronic device, and storage medium. The method obtains multiple positive sample images, multiple negative sample images, and a test sample image, obtains pixel differences based on the pixel points in the test sample image and the pixel points in the multiple positive sample images, generates a color difference threshold according to the multiple positive sample images, generates multiple feature connected regions of the test sample image according to the color difference threshold and the pixel differences, generates a first threshold according to the image noise of the multiple positive sample images, screens target regions from the multiple feature connected regions according to the number of pixel points in each feature connected region and the first threshold, generates a second threshold according to the defective pixel points of the multiple negative sample images, and determines the detection result of the test sample in the test sample image according to the area of all pixel points in the target region and the second threshold. This application can improve the accuracy of defect detection.

The invention discloses an adaptive gradient descent optimization image dead pixel compensation method and system, and relates to the technical field of image processing, and the method comprises the steps: intercepting image frames of a received image data stream collected by a short wave infrared sensor according to a preset interval; calculating a gradient descent iteration value of the image group by adopting a self-adaptive gradient descent iteration algorithm; traversing each pixel point of each image frame in the image group, pre-estimating a pre-estimated pixel value corresponding to each pixel point according to the gradient descent iteration value, and calculating a decision factor according to a difference value between the pixel value of each pixel point and the pre-estimated pixel value corresponding to the pixel value; and judging whether each pixel point is a defective pixel or not according to a comparison result of the judgment factor of each pixel point and a preset threshold value, and if the pixel point is the defective pixel, replacing the pixel value of the defective pixel with the estimated pixel value. The self-adaptive gradient descent decision factor of each pixel is calculated in real time, and the dead pixel is identified by judging the difference between the estimated value and the actual value, so that real-time, self-adaptive and high-precision identification and compensation of the dead pixel are realized.