147 results about "Quantization (image processing)" patented technology

The invention relates to the technical field of video coding and image processing, and discloses a code rate control method and system based on video image segmentation, and the method comprises the steps: carrying out the pixel-level semantic segmentation of a to-be-coded video frame sequence; extracting a foreground region of interest; calculating a corresponding segmentation uncertainty parameter; establishing semantic mutation parameters of the foreground region of interest; generating a semantic perception weight; executing region-level target code rate redistribution and quantization parameter mapping; and executing partition coding control. In the prior art, code rate control mainly depends on motion intensity or pixel complexity, and especially when a foreground target suddenly appears or disappears in a monitoring scene, a technical problem that a key target is blurred or a background code rate is wasted is easily caused. Due to the fact that the uncertainty modeling and semantic mutation sensing mechanism of the semantic segmentation result is introduced, priority coding of the foreground interest area is achieved under the frame-level code rate constraint condition, and the video coding quality and the code rate utilization efficiency are improved.

The invention relates to the technical field of point location labeling, in particular to a self-adaptive load balancing point location labeling method and system based on image processing, and the method comprises the following steps: obtaining an image source division region, extracting an edge pixel number texture direction number color channel standard deviation, generating region detail redundant structure distribution information, and monitoring node state data. The method comprises the following steps: constructing a node resource load state mapping set, matching according to a regional detail redundancy ratio and a node state, establishing a task allocation relationship, extracting a boundary gray derivative, screening a boundary stable candidate point location set, calling point location coordinates, sending the point location coordinates to corresponding nodes, executing target verification, writing in an image frame, and generating a labeling result. Quantitative extraction is performed on distribution among image region edge pixels, texture directions and color channels, and task assignment is accurately paired in combination with image structure complexity and node states, so that resource mismatching and processing retardation are avoided, and task distribution accuracy is improved.

The invention discloses an intelligent image processing method and device for an AI chip and a medium, and relates to the technical field of intelligent image processing, and the method comprises the steps: collecting original image data, carrying out K-MEANS filtering and regularization processing, and generating an image data matrix; based on the image data matrix, calculating a similarity score matrix through a query-key-value matrix, extracting an image key region, and generating a key region mask matrix and an attention weight matrix; calculating a quantization precision level parameter according to the attention weight matrix and the key area mask matrix, and performing error compensation in combination with residual learning to obtain a quantization compression feature vector; and based on the quantized compressed feature vector, performing multi-scale similarity matching with a sample library through hierarchical feature fusion, and calculating a confidence score in combination with an attention weight. According to the method, the identification precision is ensured, and the deployment efficiency of the AI chip on the edge equipment is improved.

The invention relates to the technical field of image processing, in particular to an aeronautical part identification character recognition method and system based on image processing, and the method comprises the steps: obtaining a character image of an aeronautical part, carrying out the stroke fracture reconstruction of a stroke fracture in the character image, so as to generate a reconstructed image, counting the number of connection pixel points newly added in the stroke fracture reconstruction process; performing optical character recognition on the reconstructed image to obtain a preliminary recognition result and a corresponding matching confidence coefficient; determining a stroke integrity index of the character image based on the number of the newly added connection pixel points and the total number of pixel points of the reconstructed image; the matching confidence coefficient of optical character recognition and the stroke integrity index based on reconstruction pixel quantization are combined, double judgment is carried out, it is ensured that the recognition result has the two characteristics of similar forms and reliable sources, and the accuracy and reliability of character recognition under the complex, high-light-reflection and abrasion conditions are improved.

The invention discloses an eye fundus image quality control method and device, a storage medium and electronic equipment, and relates to the technical field of image processing. The eye fundus image quality control method comprises the steps of determining a to-be-processed eye fundus image; based on an effective area of the to-be-processed eye fundus image, determining at least one type of quality quantification data corresponding to the to-be-processed eye fundus image, the effective area being used for representing an unshielded eye fundus structure area; and determining a quality control result of the to-be-processed eye fundus image based on the at least one type of quality quantification data corresponding to the to-be-processed eye fundus image, the quality control result comprising an image quality score and / or a quality problem corresponding to the image quality score. According to the eye fundus image quality control method provided by the embodiment of the invention, the score of the eye fundus image to be processed and the quality problem corresponding to the score can be intuitively known, so that related personnel can know the score and the quality problem, the normalization of the eye fundus image is improved, and standardization of eye fundus image data is facilitated.

The invention relates to the technical field of image data processing, in particular to a weld joint pseudo-defect recognition and filtering system based on image processing, which comprises the following steps: acquiring a surface image of a local area of a weld joint through an image acquisition module; marking the welding seam sub-region as a photosensitive dominant sub-region or a photosensitive non-dominant sub-region through a region marking module; screening defect confidence suspected sub-regions through an initial recognition module according to a comparison result of texture feature quantized values of the photosensitive dominant sub-regions and the photosensitive non-dominant sub-regions; analyzing the spatial distribution condition of the defect confidence suspected sub-regions through a feature matching module so as to judge the defect filtering type of the local region of the welding seam; and determining a defect filtering strategy for the local area of the welding seam through an identification filtering module. Furthermore, by using the difference between the pseudo defect and the real defect in the optical response rule, a targeted regional differentiation analysis mechanism is adopted, and the pertinence of the filtering strategy is improved.

The application discloses a kind of based on data center application inference big model discovers network topology fault and false alarm self-healing visualization method and system thereof, including using grid division algorithm carries out grid division and forms DeepSeek inference network topology and chip resource mark;Visual image processing is carried out in DeepSeek inference network topology, complete network topology resource re-planning and visual data annotation, output DeepSeek inference big model inference conclusion;From network topology gateway configuration file, the transmission protocol of application is used as input parameter and carried out Deepseek inference big model inference analysis, carries out power quantization;DeepSeek inference network topology node is associated, and alarm is carried out in transmission protocol through the fluctuation of power level.The application uses microfluidic chip to manage the chip that inference big model carries out operation, chip resource mark is input into inference big model, and then uses image definition position and inference big model embedded AI algorithm matching discovers network topology, improves big model inference conclusion and visual accuracy.

The invention relates to the technical field of image processing, and discloses an AI image super-resolution reconstruction method based on a multi-scale fusion mechanism. The AI image super-resolution reconstruction method based on the multi-scale fusion mechanism comprises the following steps: establishing a binocular image system; establishing a double-end collaborative model architecture; establishing a front-end model optimization mechanism; according to the method, through a scene adaptive feature labeling decision model, three-dimensional position driven quantization and spectral response calibration replace manual presetting, and parallax-pose linkage correction is combined, so that subjective interference is thoroughly eliminated, and the misalign error suppression effect is improved by more than 40%; scale intelligent selection and cross-scale feature interaction are introduced into a dynamic interactive multi-scale attention convolutional neural network (DI-MSCNN), and dynamic definition of convolution kernel parameters is matched, so that feature extraction efficiency is improved by 30%-40%, and texture density differences can be accurately adapted; a detail hierarchical perception GAN (DLP-GAN) generates a strategy through hierarchical discrimination and detail partitioning.

The application discloses a kind of crack segmentation image post-processing optimization and quantization method, device and equipment, it is related to computer vision and digital image processing technical field, this method includes obtaining the initial binary crack mask chart corresponding to the image to be processed, and carries out geometric morphological filtering operation to eliminate non-crack area, obtains preliminary filtered mask chart;Contour detection is carried out to the preliminary filtered mask chart, and the crack-containing region of interest mask chart is identified and cropped;Break point connection processing is carried out to the region of interest mask chart to repair discontinuous crack fragment, generate continuous optimized crack mask chart, to carry out the geometric size parameter calculation of crack.This application can significantly improve the accuracy, integrity of crack segmentation result, and can realize the automation, accurate quantization of key parameters.

The application relates to the field of image processing and discloses a bone metastasis detection and positioning system and method based on SPECT bone imaging. The system extracts and processes multi-scale features of a to-be-detected image through a feature pyramid network to obtain a multi-scale feature map; then extracts metabolic features, morphological features and uptake mode features from the multi-scale feature map through a bone-specific feature model, and gradually fuses the metabolic features, the morphological features and the uptake mode features to obtain a bone-specific enhanced feature map; then extracts and fuses features of the bone-specific enhanced feature map through a mixed feature fusion model to obtain mixed fusion features; finally, the mixed fusion features are subjected to probability reasoning and double-uncertainty quantization processing through a probability reasoning network to determine a bone metastasis detection result, high-precision detection and positioning of bone metastasis lesions are realized, and the problems that existing technologies cannot effectively identify different size lesions and lack global perception ability are effectively solved.

The application provides a model quantification method, device and equipment and storage medium, the method comprises: using integer algorithm to execute a plurality of operators of transformer model for image processing, the plurality of operators of the transformer model comprises linear operator and nonlinear operator, wherein the linear operator comprises matrix multiplication operator, the matrix multiplication operator adopts symmetric quantization to quantize floating point value to integer value;The nonlinear operator comprises an activation function operator and a layer normalization operator, the activation function operator adopts polynomial fitting to quantize floating point value to integer value, and the layer normalization operator adopts the mean and standard deviation of input data in the channel dimension to quantize floating point value to integer value.The transformer operator is quantified, so that the inference work of the transformer model for natural language is based on integer operation, so that it can be truly deployed on FPAG chip for practical application.

This invention relates to the field of medical image processing, specifically to a skin scar image segmentation method for automated skin scar region identification and quantitative analysis. It addresses the problems of low segmentation accuracy, inaccurate boundary modeling, and unreliable quantization calculations caused by insufficient feature representation, weak generalization ability, and reliance on manual intervention in existing methods. The method includes: S1 acquiring a skin scar image; S2 inputting a trained improved SAM2 segmentation network; and S3 outputting a segmentation mask. The improved SAM2 network is trained by constructing and preprocessing a dataset; it includes a frozen-parameter backbone image encoder, a cue encoder, a medical image feature guidance module (extracting texture details, color contrast, and edge morphology features), a residual enhancement module (fusing general features, cue vectors, and medical features through residual connections), and a mask decoder. During training, the backbone network parameters are frozen, and only the cue encoder, medical feature guidance module, residual enhancement module, and mask decoder are fine-tuned.

The invention provides a machine vision coding method based on feature distillation, and relates to the technical field of image processing.The method comprises the steps that an image to be processed is input into a machine vision coding model, and the model extracts first potential feature representation of multiple channels through an analysis encoder; the method comprises the following steps: quantitatively dividing into basic layer quantitative features containing semantic and spatial structure features and enhancement layer quantitative features containing detail and texture features; the hyper-priori correlation module encodes hyper-priori information and generates enhanced auxiliary features and basic auxiliary features, and the conditional entropy coding network realizes encoding and decoding of the basic layer quantization features and the enhanced layer quantization features based on the enhanced auxiliary features and the basic auxiliary features. A machine vision task result is obtained through a feature transformation and task processing module, and after splicing is conducted through a splicer, a reconstructed image is output through a synthesis decoder. According to the invention, both machine vision and image reconstruction can be considered.

An image processing apparatus for forming an enhanced image is disclosed. The apparatus comprises one or more processors configured to: receive one or more input images; form, from each of the one or more input images, a respective feature representation, each feature representation representing features of the respective input image; and subject the one or more feature representations to a symmetric pooling operation to form an enhanced image from at least some of the features of the one or more feature representations identified by the symmetric pooling operation. The apparatus may generate images with increased photoreceptive dynamic range, increased bit depth and signal-to-noise ratio, with less quantization error and richer colour representation.

The embodiments of the present specification provide an image encoding method, an image decoding method and an image processing system, wherein the image encoding method comprises: obtaining an original residual of a first image encoding unit, wherein the original residual is used to reflect the difference between the original pixel and the predicted pixel of the first image encoding unit; determining a unit index of the first image encoding unit based on the original residual, wherein the unit index is used to reflect the content complexity of the first image encoding unit; and performing quantization encoding on a second image encoding unit based on the unit index to obtain an image encoding result, wherein the first image encoding unit and the second image encoding unit are image encoding units obtained by dividing the same image at different encoding stages. By using the unit index to guide the quantization encoding, the quantization precision can be dynamically adjusted during the encoding process. Finally, while maintaining the subjective visual quality, the code rate allocation is optimized, and the overall encoding performance and compression efficiency are improved.

The application relates to an NPU-based deep learning model training and deployment optimization method, relates to the technical field of image processing, and comprises the following steps: constructing a deep learning model and training; based on a comprehensive quantization precision optimization algorithm and sample images corresponding to multiple target detection tasks, determining an optimal quantization precision strategy corresponding to each sample target detection task, combining sample images corresponding to target detection tasks of a currently deployed edge device, and configuring the quantization precision of a pre-trained deep learning model deployed to the edge device; based on sample images corresponding to target detection tasks of the currently deployed edge device and a dynamic threshold adjustment mechanism corresponding to each sample target detection task, constructing the dynamic threshold adjustment mechanism corresponding to the target detection tasks of the currently deployed edge device, embedding the configured pre-trained deep learning model, and deploying the deep learning model, so that the effect of deploying the deep learning model on the edge device is improved.

This invention discloses a method and system for candidate optimization, continuous repair, and quantization of crack images in complex backgrounds, belonging to the field of computer vision and digital image processing technology. To address the problems of crack recognition in complex backgrounds being susceptible to false target interference, discontinuous cracks, and difficulty in stable quantification of engineering parameters, this invention obtains the crack response probability map and determines a linked dual threshold. It retains low-threshold regions connected to the high-threshold region to generate initial candidate regions. The ratio of skeleton length to area is calculated as a thinness index for screening. Skeleton endpoints are extracted, and bridging pixel paths are generated to achieve local repair when the endpoints of different connected domains meet preset multi-condition constraints. The repaired region is then skeletonized to construct a graph structure. The number of original skeleton segments, the total number of cracks, and the total length are statistically analyzed. Width statistics and direction parameters are calculated by sampling along the skeleton using distance transformation. This effectively suppresses false targets in complex backgrounds, accurately connects discontinuous cracks, and outputs stable, accurate, and interpretable engineering parameters.

This application provides an image encoding and compression method, apparatus, storage medium, and program product. The method relates to the field of image processing. Before image encoding, it extracts the content features of the original image, obtains at least one distorted image of the original image, and extracts the content features and distortion behavior features of each distorted image. Using a prediction model, based on the quality target, the content features of the original image, and the content features and distortion behavior features of each distorted image, it predicts the quantization step size corresponding to the quality target. The original image is then encoded according to the corresponding quantization step size. By fitting the relationship between the quantization behavior quantization step size and the quality index through the prediction model, the corresponding quantization step size can be inversely derived from the quality target of the encoding. Furthermore, the image encoding quality can be precisely controlled according to the corresponding quantization step size, improving the subjective quality of the image encoding result and ensuring that the image encoding result meets the given quality target.

The invention relates to a non-standard JPEG (Joint Photographic Experts Group) coding processing method for dynamic code rate control. The method comprises the following steps: presetting a multi-frame joint image processing scene, obtaining an original reference image, and carrying out combined processing of boundary expansion and random bias addition to obtain a preprocessed reference image; dividing macro blocks for the preprocessed reference image, and distributing an independent coding parameter configuration space after verification and compliance; frequency domain conversion is executed according to a division result, an initial quantization parameter is calculated through an index model, and a dynamic quantization table is dynamically adjusted; executing non-standard entropy coding to obtain an indexed macro block coding code stream; combining the indexed code stream and a macro block division result to construct an index table containing a macro block serial number, an initial address and a code stream length; analyzing the code stream to obtain quantized data, and performing inverse quantization by using a dynamic quantization table to obtain spatial domain data; enabling the format of the reconstructed image to be consistent with that of the preprocessed reference image through reverse preprocessing, and performing real-time processing on the linkage input image to obtain a reconstructed reference image; and non-standard JPEG coding processing of dynamic code rate control is realized.

The utility model relates to a special X-ray detection detector for a building, which comprises a cylindrical closed shell, a detector body and a detector body. The detector element is arranged in the cylindrical closed shell and is used for completing X-ray sensing, pixel selection, signal quantization and image processing in the axial direction of the cylindrical closed shell and outputting digital image data; the angle indication module is arranged at the end part of the cylindrical closed shell and is used for indicating the current angle of the detector so as to be conveniently aligned with a ray incident angle; and the waterproof and dustproof connector is arranged at the end part of the cylindrical closed shell, is electrically connected with the detector element, and is used for power supply and data interaction. The device can be matched with an existing building prefabricated hole channel, wired stable transmission and rapid angle positioning can be achieved, and high-efficiency and low-damage ray detection of the interior of a bearing component in the true sense is achieved.

The application provides a kind of method for enhancing robustness of pulse neural network based on neural activation and connection optimization, applied to artificial intelligence and neural network technical field, the method comprises: constructing pulse neural network model, the neuron of pulse neural network model adopts burst enhancement type pulse neuron, the input of pulse neural network model is image data, and the output of pulse neural network model is the image processing result corresponding to computer vision task;When the membrane potential of burst enhancement type pulse neuron exceeds the firing threshold, the part exceeding the membrane potential is converted into the pulse output within the burst window by quantization linear mapping function, and the number of pulse output is an integer between 0 and the preset maximum burst pulse number;When training pulse neural network model, add activation perception regularization term to total loss function;Through the application, the accuracy and robustness can be simultaneously improved while maintaining the high energy efficiency advantage of pulse neural network.

The invention discloses a real-time polyp segmentation system based on a multi-domain hierarchical attention network, and relates to the field of medical image processing and computer vision, and the system comprises a dynamic region guide block which is used for partitioning an input image and selecting key region features through a two-stage routing attention mechanism; the potential entropy quantization channel space attention module is used for carrying out channel and space information entropy quantization calculation on the feature map and highlighting fine organization differences; the spatial frequency fusion module is used for performing pixel-level fusion on the spatial domain features and frequency domain features extracted through fast Fourier convolution, and capturing a periodic mode and global structure information; and the patch extension layer and the linear projection are used for layer-by-layer up-sampling and executing quadruple up-sampling in the final stage to recover to the input resolution, and pixel-level segmentation prediction is generated. The boundary precision and the pixel classification performance are both improved, and the real-time performance of polyp segmentation and the clinical application reliability are remarkably improved.

The present application relates to the technical field of image data processing, and more particularly to a welding seam pseudo-defect identification and filtering system based on image processing, which acquires a surface image of a local area of a welding seam through an image acquisition module; labels a sub-area of the welding seam as a photosensitive dominant sub-area or a photosensitive non-dominant sub-area through a region labeling module; screens a defect confidence suspected sub-area according to a comparison result of texture characteristic quantization values of the photosensitive dominant sub-area and the photosensitive non-dominant sub-area through a preliminary identification module; analyzes a spatial distribution of the defect confidence suspected sub-area through a feature matching module to determine a defect filtering type of the local area of the welding seam; and determines a defect filtering strategy for the local area of the welding seam through an identification and filtering module. Thus, the difference between pseudo-defects and real defects in optical response rules is utilized, a targeted area differentiation analysis mechanism is adopted, and the targeting of the filtering strategy is improved.

The application discloses a kind of high-efficiency Gaussian splashing reconstruction method and device for dynamic endoscope scene in the technical field of medical image processing, comprising: the endoscope video sequence obtained is preprocessed, and initial Gaussian point cloud is generated;The motion trajectory of each Gaussian point in initial Gaussian point cloud is modeled using Gaussian point cloud morphing model based on discrete cosine transform, and dynamic Gaussian scene is generated;The dynamic and static attributes of Gaussian point cloud morphing model based on discrete cosine transform are compressed and stored using residual-aware hybrid precision quantization strategy;Dynamic Gaussian scene is rendered using hardware-aware dense inference strategy, and three-dimensional reconstruction image of endoscope video is generated in real time.The application can effectively improve the expression efficiency of endoscope scene dynamic modeling, while reducing storage / deployment overhead.

The invention relates to the technical field of synthetic aperture radar image processing, and discloses an SAR image quantization and enhancement method based on histogram statistics, and the method comprises the steps: obtaining a minimum value and a maximum value of an SAR amplitude diagram; setting parameters required for SAR image quantization and enhancement, performing normalization processing on the SAR amplitude diagram according to a histogram statistical order to obtain a first normalized amplitude diagram, and extracting sample points from the first normalized amplitude diagram; determining an amplitude quantization upper limit according to the histogram result and the image brightness adjustment factor; according to the amplitude quantization upper limit, performing amplitude limiting processing on the first normalized amplitude diagram, and performing second normalization to obtain a second normalized amplitude diagram; performing contrast enhancement on the second normalized amplitude diagram to obtain an enhanced amplitude diagram; the SAR image enhancement method and the SAR image enhancement device solve the technical problems that according to an existing enhancement method, noise is prone to being amplified, brightness distortion is caused, or contrast ratio improvement is insufficient, and the overall processing efficiency is low.

The invention discloses a color correction method and system, and belongs to the technical field of image processing, and the method comprises the steps: extracting a chromaticity quantization DCT coefficient, an initial chromaticity quantization table and a standard color card calibration matrix of a compressed image; wherein the chroma quantization DCT coefficient comprises a DCT coefficient of a blue chroma component and a DCT coefficient of a red chroma component; generating a chromaticity correction matrix according to the DCT coefficient of the blue chromaticity component of the compressed image, the DCT coefficient of the red chromaticity component of the compressed image and the standard color card calibration matrix; correcting the frequency domain chromaticity of the compressed image according to the chromaticity correction matrix; generating a final chroma quantization DCT coefficient according to the initial chroma quantization table and the chroma correction matrix; outputting a compressed image code stream according to the final chromaticity quantization DCT coefficient and the entropy coding table; according to the method, chromaticity correction is directly carried out in the frequency domain of the compressed image, quantization table optimization and chromaticity statistical feature modeling are combined, the problems of calculation complexity and distortion caused by decompression processing are avoided, and the color fidelity of the compressed image is improved.

The invention provides an image compression method and device, electronic equipment and a readable medium, and relates to the technical field of image processing, and the image compression method comprises the steps: obtaining an input image and a target code rate; inputting the input image into the content feature analysis model to input a feature vector of the image, the feature vector being capable of reflecting a perception feature of the input image; and inputting the target code rate and the feature vector into a joint parameter decision model to generate a spatial scaling parameter and a quantization parameter adaptive to the input image and the target code rate. According to the embodiment of the invention, under the limitation of the target code rate, the details of the visual salient region are reserved preferentially according to the adaptive control of the image prediction optimal compression parameter, and the method is widely applied to application scenes such as extremely-low bandwidth image transmission, terminal front-end compression and Internet of Things image perception.

The invention discloses a visual detection acceleration method based on model quantification, belongs to the technical field of visual detection, is used for visual detection of cone yarns, and comprises the steps of obtaining enterprise historical data, performing data cleaning, data integration, data transformation and data specification, and performing feature selection, feature construction and feature fusion in combination with a database. And importing a feature fusion result into a resultant yarn quality detection model in combination with domain knowledge, and performing model evaluation, comparison and optimization. By integrating the high-resolution industrial camera and an advanced image processing algorithm, non-contact and high-precision analysis of the appearance state of the cone yarn is realized. According to the technology, key information such as cone yarn winding forms, surface flaws and yarn breakage can be captured in real time, and the automation level of the spinning production process is remarkably improved.

This application provides an image compression system, an image processing method, an encoding / decoding method, and an electronic device. The image compression system includes a first selection module, an entropy encoding module, an entropy decoding module, a quantization module, N encoding networks, and one decoding network. The N encoding networks have different encoding losses. The first selection module is used to select a target encoding network from the N encoding networks based on the number of times the image to be encoded has been encoded. The target encoding network is used to perform feature transformation on the image to be encoded to obtain a first feature map. The quantization module is used to quantize the first feature map to obtain a second feature map. The entropy encoding module is used to entropy encode the second feature map to obtain a bitstream. The entropy decoding module is used to entropy decode the bitstream to obtain a third feature map. The decoding network is used to perform feature transformation based on the third feature map to obtain a reconstructed image. This effectively reduces the loss ratio of the image after multiple encoding and decoding operations.

The invention provides a method and system for converting an RGB image into a depth map based on a variational auto-encoder, and relates to the field of image processing, and the method comprises the steps: constructing and training an image conversion model which comprises a feature extraction network, a probability encoder, a self-adaptive sampling unit, a condition decoder and an uncertainty quantization unit, the feature extraction network is used for extracting a feature map of the RGB image, the probability encoder is used for carrying out feature extraction on the feature map and outputting a mean value and a variance of a potential space, and the adaptive sampling unit is used for extracting noise influence features from the RGB image, determining a plurality of sampling noises and outputting the sampling noises to the RGB image. The condition decoder is used for generating a plurality of hypothetical depth maps corresponding to the RGB image based on the mean value and variance of the potential space and the plurality of sampling noises so as to generate a depth map, and the uncertainty quantization unit is used for generating an uncertainty map corresponding to the RGB image; the depth map corresponding to the to-be-converted RGB image is generated through the image conversion model. The method has the advantage of improving the reliability of depth map conversion.