306 results about "Structural similarity" patented technology

The invention discloses a microscopic image seamless splicing and enhanced reconstruction method based on deep learning, and the method comprises the following steps: S1, collecting a plurality of original images with overlapped regions, and recording the spatial position information and imaging parameters of the original images; s2, preprocessing the original image to generate a standardized image sequence; s3, inputting the standardized image into a structure perception feature extraction network, and extracting a feature map fusing textures and structures; s4, inputting the feature image and the original image into an image registration module; s5, inputting the registration image into the boundary attention splicing network; s6, inputting the seamless image into the residual hierarchy reconstruction network, and enhancing image details through hole convolution and multi-scale branches; s7, image quality evaluation is executed, and the structural similarity, the signal-to-noise ratio and the edge retention rate are calculated; and S8, constructing a training set and carrying out end-to-end training optimization based on a joint loss function. According to the method, a multi-module deep network is fused, and seamless splicing and high-quality enhanced reconstruction of microscopic images are realized.

The invention discloses an MRI-PET mode conversion method and system based on a generative adversarial network, and belongs to the technical field of artificial intelligence medical image generation. And the multi-scale structure representation injection module injects multi-scale anatomical prior information at different stages of the encoder, and overcomes the limitations of insufficient utilization of prior information and single injection scale. And the adaptive semantic residual fusion module adopts semantic attention guidance and double-branch attention weighting, adaptively fuses fine-grained local features and global context information, harmonizes the difference between the fine-grained local features and the global context information in an abstract level and a semantic category, and solves the problems of feature conflict and semantic fuzziness in a bottleneck region. The direction sensing space-frequency discriminator realizes multi-dimensional and fine-grained adversarial supervision through a space, frequency and local image block multi-branch collaborative discrimination mechanism, and improves the structural fidelity and spectrum authenticity of a synthetic image. And the generated image is superior to the existing method in indexes such as structural similarity and peak signal-to-noise ratio, and has higher clinical practical value.

The invention relates to the technical field of three-dimensional reconstruction, in particular to a near-infrared assisted low-light scene three-dimensional reconstruction method based on 3D Gaussian splashing, which is characterized in that generation of a Gaussian ellipsoid is dominated by a near-infrared image with a high signal-to-noise ratio, and a stable geometric basis is provided for a visible light information missing area; in the rendering stage, a near-infrared rendering image and a normal visible light rendering image are respectively generated through Gaussian ellipsoid shared geometric parameters and respective opacity and color attributes of two modes; through cross-modal structure similarity loss, a normal visible light image which is forcibly rendered is aligned with a near-infrared light image in structure, and clear near-infrared structure information is used to strictly constrain the recovery process of the color of the visible light image, so that the accuracy and authenticity of a recovery result are ensured.

The invention discloses a small sample remote sensing image classification method based on hierarchical spatial structure learning. The method comprises the following steps: firstly, extracting multi-scale features of a remote sensing image by using a ViT (Visual Transform) model, and capturing rich semantic information and spatial structure relationships in the image; secondly, constructing a graph structure based on spatial adjacency and attention weight to model a structured relationship between samples, and encoding graph node features through a graph convolutional network (GCN) so as to enhance the discrimination ability of the features in a structural semantic space; thirdly, a residual enhancement mechanism is introduced to fuse global semantic information, and the discrimination capability of graph embedding is improved; then, based on the structural similarity between the support set and the query set, performing classification decision, and realizing accurate classification under a small sample condition; and finally, carrying out joint optimization on the whole model by adopting a training strategy of a small sample meta learning task and a supervision loss function.

The invention discloses a small sample cable protection layer overfire temperature-image generation method based on meta-learning, and the method comprises the steps: constructing a multi-task small sample data set through a small number of fire resistance test images, proposing a meta-learning type image generation model, and achieving the quick adaption to different protection systems through an internal and external circulation meta-learning strategy; inputting a target protection configuration, a fire condition and a temperature condition, wherein the meta-learning type image generation model can generate a fire appearance image of each protection layer at a corresponding temperature grade; further performing physical consistency judgment based on temperature-damage trend, cross-layer association and protection system configuration to ensure that a generation result conforms to a real damage rule, and performing quantitative index verification by calculating a structural similarity index, learning and sensing image block similarity and a Frechet Inception distance; according to the method, the multi-working-condition high-quality fire passing image can be generated under the condition of image scarcity, the fire resistance test cost is reduced, and data support is provided for temperature inversion and damage evaluation after a bridge cable fire disaster.

The invention relates to the cross technical field of artificial intelligence and medical image processing, in particular to a medical image segmentation system and method based on a wavelet bridge diffusion model and an efficient conditional random field. A WBDM-ECRF framework is constructed and comprises a discrete wavelet transform module, a BDM-T module, a BDM-S module and an ECRF module; decomposing the image through discrete wavelet transform, extracting a low-frequency sub-band, and enhancing the contrast ratio of a focus and normal tissues; the BDM-T takes U-Net as a backbone, integrates a FlashAttention mechanism, and optimizes a variance formula to realize efficient training; the BDM-S adopts a leapfrog sampling strategy, so that the reasoning time is greatly shortened; the ECRF introduces a multivariate potential function of a structural similarity index and smooth operation through edge expansion, and accurately optimizes edge segmentation. According to the method, the dependence of marked data is reduced, the training and reasoning efficiency is improved, the problem of fuzzy edge segmentation is solved, the Dice coefficient and intersection-union ratio performance on the ISIC data set is excellent, and reliable quantitative support is provided for disease diagnosis and treatment.

The invention discloses a wind turbine generator impeller anomaly detection method and system based on sound vibration signal identification. According to the method, firstly, impeller response is collected through a microphone / vibration sensor, and a Mel spectrogram is generated; then, a Teager-Kaiser energy operator and self-correlation analysis are utilized, and under the condition of not depending on a rotating speed signal, the rotating period of the impeller is recognized, and frequency spectrum segmentation is carried out; calculating a cross-period Mel frequency spectrum dynamic deviation, and normalizing the cross-period Mel frequency spectrum dynamic deviation through an amplitude correction coefficient related to the rotating speed; generating a periodic coherent energy diagram by adopting an improved structural similarity algorithm; performing filtering enhancement by using a harmonic resonance template, performing morphological deconstruction and parameterization on an abnormal region in the graph, and extracting geometric features; and finally, calculating a comprehensive abnormal score based on the multi-dimensional features and realizing automatic early warning. According to the method, the problems of variable working condition interference and rotating speed dependence are effectively solved, and accurate and stable detection of early abnormality of the impeller can be realized.

A multi-feature three-dimensional reconstruction 3D Gaussian method based on laser vision comprises the steps that laser radar point cloud and camera images are aligned through space-time calibration, and a unified coordinate system is established; extracting geometric features by using point cloud data acquired by the Lidar point cloud, and initializing a Gaussian ellipsoid according to the Lidar point cloud; optimizing the brightness, the contrast ratio and the structural similarity of the rendered image and the real image by combining the mean absolute error L1 and the structural similarity SSIM; the curvatures of Gaussian ellipsoids of K-nearest neighbors are forced to be consistent, and long and short axes and line and surface features of the Gaussian ellipsoids are aligned to reduce geometric distortion; the distribution density of 3D Gaussian is dynamically adjusted through line / surface features and visual structure information extracted by Lidar, and balance between geometric detail enhancement and calculation efficiency is achieved. According to the method, the position, the scale and the rotation parameters of Gaussian are uniformly optimized, and the details and the calculation efficiency of the model are balanced while the consistency of the model structure is improved.

The invention belongs to the technical field of charging gun detection, and discloses a charging gun damage detection method and computer equipment, and the method comprises the steps: obtaining a detection picture of a charging gun of which the damage condition is to be detected; calculating a structural similarity index SSIM of the detection picture and a preset standard picture, and drawing a damaged area, which is different from the standard picture, on the detection picture; calculating a coverage value of the damaged area covering a preset key area and a damage proportion value of the area of the damaged area in the whole area of the detection picture; performing preset weighted calculation based on the coverage value and the damage proportion value to obtain a damage evaluation value of the charging gun; and generating corresponding alarm information based on the damage evaluation value. Through the multi-dimensional damage assessment system, the problem that the misjudgment rate is high due to the fact that damage is judged only through pixel differences in the prior art is effectively solved.

The invention discloses a multi-mode ultrasonic microscopic image contrast enhancement method fused with expert priori knowledge, and the contrast, definition and detection reliability of an ultrasonic microscopic image are improved. The method comprises the following steps: a multi-modal feature coding module based on a CLIP framework extracts feature representations of an ultrasonic microscopic image and an expert cue word, and maps the feature representations to a unified semantic space through a cross-modal alignment mechanism; designing a semantic guidance prompt module, constructing semantic elements which highlight detection demand guidance, and combining an image-quality description sample to carry out few-sample fine tuning so as to enhance the response capability of the model to specific semantics; an image enhancement and reconstruction module is designed, a coding-decoding structure and a cross-layer feature connection mechanism are adopted, contrast enhancement, detail reconstruction and noise suppression are realized under the guidance of a CLIP semantic vector, self-learning correction is carried out based on a standard grooving plate to improve contrast performance, and optimization training is carried out in combination with structural similarity loss and semantic consistency loss.

The invention discloses a music database retrieval method and system based on feature extraction, and relates to the technical field of data analysis, and the method comprises the steps: carrying out the preprocessing of an audio signal of a query track input by a user, dividing the audio into equal-length frame sequences, and generating three-channel spectrogram tensors, including a Mel spectrogram, a logarithmic amplitude spectrogram and a CQT spectrogram; extracting multi-modal features of the equal-length frame sequence, dividing melody motivation segments according to pitch change and stability of the audio signal, performing differential coding, generating a melody differential sequence, and modeling the melody differential sequence into a melody topological graph; and calculating the structural similarity between the melody topological graph and a database melody graph, screening a similar melody candidate set P, mapping a melody curve into a topological manifold through a topological data analysis method, extracting persistent homology features, and converting the persistent homology features into a persistent bar graph. According to the method, the retrieval precision and the matching credibility of the complex melody in the music database are remarkably improved.

The invention relates to an abnormal event detection method and device, an electronic device and a storage medium, and the method comprises the steps: obtaining to-be-detected video data in a target detection scene; based on 3D-CNN, spatio-temporal feature extraction is carried out on the to-be-detected video data to obtain a spatio-temporal feature sequence; inputting the spatial-temporal feature sequence into a Transform encoder to obtain encoding information of each video frame; performing future frame prediction according to the coding information of each video frame by using a diffusion model to obtain future frame information; and performing structural similarity comparison on the future frame information and video frame information in the to-be-detected video data, and determining an abnormal event detection result of the target detection scene based on a comparison result. Based on the combination of a 3D-CNN network, a Transform and a diffusion model, multi-scale prediction is realized, scenes in different time periods are covered, and the accuracy of tracking and predicting a target motion trajectory in video data is improved.

The invention discloses a student answer text similarity detection method for artificial intelligence paper marking. The method comprises the following steps: converting a target answer sheet into an electronic text; according to the comprehensive similarity between the student answering text and the question stem text in the electronic text, identifying and positioning similar specific sentences or fragments; comparing the paragraph-level semantic similarity, the key fragment dynamic score, the structural similarity and the keyword coincidence rate according to the student answer text and a model essay text in a preset model essay library, and generating a comprehensive similar risk score; according to a preset multi-level early warning threshold value, outputting a corresponding similar risk early warning level; and presenting the specific sentences or fragments and the same risk early warning level. Therefore, through fusion of question stem similar positioning and model essay multi-dimensional comparison, a comprehensive similar risk score of graded early warning is generated, the transformation from manual low-efficiency screening to intelligent accurate identification is realized, and the paper marking efficiency, the similar detection comprehensiveness and the scoring fairness are remarkably improved.

The invention discloses a neural network reconstruction method for a multispectral image, which is used for performing complete band sequence reconstruction on an acquired limited band image, and comprises the following steps: S11, inputting a multi-channel image into a trained neural network; s21, feature extraction and multi-scale coding are carried out; s31, the reconstruction reasoning module outputs all target wave bands; and S41, comparing with a real image and calculating loss, wherein the loss comprises MSE, frequency domain residual error and structural similarity. The invention provides a neural network reconstruction module with wave band prediction capability, which is specially used for recovering a complete wave band image sequence from a few physically acquired key wave band images.

The invention discloses a laser shearing speckle interference phase unwrapping method based on deep learning. According to the method, eight speckle patterns before and after deformation are used as input, multi-scale feature extraction is carried out through layer-by-layer convolution and down-sampling of an encoder, features are sent to an attention fusion module, response weights of channels and spatial positions are adaptively adjusted through channel attention and position attention, then the response weights are input into a decoder, and jump connection with an encoding end is combined, so that multi-scale feature fusion is realized. And step-by-step reconstruction of multi-scale features is realized, and a phase diagram is output. Performing comprehensive constraint on prediction and reference phases in the aspects of numerical deviation, structural consistency and gradient smoothness and updating network parameters by adopting composite loss formed by mean square error, mean absolute error, structural similarity, gradient loss and out-of-plane displacement calculation items; therefore, an unwrapping result which is globally continuous and has clear and stable local details is obtained, and a feasible way is provided for intelligent processing and automatic analysis of the laser shearing speckle interference image.

The invention discloses an event camera image reconstruction method of a multi-frame fusion network based on optical flow guidance, and belongs to the technical field of computer vision and image processing. The method comprises the following steps of: 1, acquiring event data, and converting the event data into continuous and smooth space-time voxels by adopting a full-time interval voxel coding scheme based on Gaussian distribution; step 2, constructing an FMF-Net network guided by an optical flow; step 3, designing a comprehensive loss function used for training the FMF-Net network; 4, performing supervised training on the FMF-Net network by using the public event data set; and step 5, inputting space-time voxels by using the trained FMF-Net network, and outputting a high-fidelity reconstructed image. According to the method, long-time motion clues and time domain dynamic features can be fully utilized, high-fidelity image reconstruction is realized under high-speed motion and sparse event input, the problems of loss of reconstruction details and poor consistency of an existing method in a dynamic scene are solved, and the structural similarity and visual quality of a reconstructed image are remarkably improved.

The invention relates to the technical field of document duplicate checking. The document duplicate checking method comprises the steps that according to syntactic dependency structures in text information of a document to be subjected to duplicate checking and text information of a plurality of comparison documents, keywords and context dependency relationships of each sentence are recognized, a semantic map is constructed, and the semantic map is subjected to duplicate checking; mapping the keywords and the context dependency relationship to corresponding concept nodes in the semantic graph, constructing a semantic sub-graph of the document to be subjected to duplicate checking and a plurality of semantic sub-graphs of the comparison document, and comparing the structural similarity between the semantic sub-graph of the document to be subjected to duplicate checking and the semantic sub-graphs of the comparison document to obtain a structural similarity score; and based on a node sequence in the semantic subgraph of the document to be subjected to duplicate checking, constructing a disturbance semantic subgraph, calculating a disturbance score fluctuation range, screening a comparison document of which the structural similarity score is higher than a preset threshold value and the disturbance score fluctuation range is lower than a disturbance tolerance threshold value, and outputting duplicate checking difference information. The method and the device have the effect of improving the document duplicate checking accuracy.

The invention provides an industrial defect image generation method based on conditional diffusion and multi-dimensional feature collaborative optimization, and belongs to the technical field of image processing. Comprising the steps of defining overall input and output of a system; constructing a few-sample feature extractor based on self-supervised learning; carrying out condition vector construction and feature fusion; optimizing a multi-loss collaborative conditional diffusion generator; reasoning generation and high-quality sample screening; and outputting and evaluating the quality. The method comprises the following steps: constructing a multi-modal condition vector fusing a reference defect image, a normal object image and a target area mask; and a composite feature learning loss function composed of structural similarity, perceptual hash, gradient and semantic consistency is introduced, a self-supervised contrast learning pre-training feature extractor is adopted, and automatic sample screening is performed by using a maximum edge correlation algorithm after generation. According to the method, the industrial defect image which is highly vivid, rich in details and provided with pixel-level accurate labeling can be generated from a very small number of reference images.

The invention relates to the technical field of remote sensing image processing, in particular to a homomorphic filtering-CLAHE remote sensing image enhancement method based on integration strategy multi-target particle swarm optimization, which comprises the following steps: acquiring remote sensing image data, and preprocessing the remote sensing image data; converting the preprocessed remote sensing image into an HSV color space, and extracting a V component in the HSV color space; constructing a frequency domain-spatial domain hybrid enhancement framework of homomorphic filtering and contrast-limited adaptive histogram equalization, and performing enhancement processing on a V-channel component by using the framework; constructing an integrated strategy multi-target particle swarm optimization algorithm; constructing a four-target fitness function including structural similarity, average gradient, information entropy and gray variance, and guiding particles to search in the direction of optimizing a plurality of key image quality indexes at the same time by using the function so as to realize optimal selection of remote sensing image enhancement parameters; the method can effectively enhance the definition and structural integrity of the terrain texture in the remote sensing image of the complex mountainous area.

The invention discloses a method and a system for automatically repairing context learning code vulnerabilities of a large language model, and aims to solve the problems that invalid patches are easily generated, vulnerability causes are misunderstood and verification is lacked in existing LLM zero sample repairing, and a traditional method depends on annotated data or test input. The method comprises the following steps: processing a vulnerability code-patch pair containing a plurality of different CWE types and a target vulnerability code, extracting vulnerability-related codes and constructing an example pool; an adaptive example is selected through comprehensive comparison of semantics, lexical and structural similarity; generating prompt words in combination with related codes and adaptive examples, driving LLMs to generate candidate patches, and outputting effective patches after verification of multiple LLMs; the system comprises a code processing module, an example selection module and a patch generation and verification module. According to the method, LLM and a large amount of annotated data do not need to be finely adjusted, the repair efficiency and accuracy are improved, the labor cost is reduced, multi-language extension is supported, and industrialization potential is achieved.

The invention discloses an electrocardiogram technology synthesized by an auto-encoder, relates to the technical field of crossing of medical image processing and computer vision, in particular to the electrocardiogram technology synthesized by the auto-encoder, and aims to solve the problem that an existing ECG image generation method is difficult to simulate real paper wrinkles and geometric deformation. According to the method, ECG images are input in a partitioned mode to serve as sub-blocks, pixel offset is extracted from style images for geometric deformation, a diffusion model is adopted for style transfer, content feature anchoring and a style mapping enhancement mechanism are combined, and the vivid wrinkle visual effect is injected while it is ensured that ECG waveform features are reserved. The image generated by the method has high structural similarity and visual authenticity, is suitable for medical data enhancement and diagnosis model training, and improves the generalization ability and clinical practicability of downstream tasks.

This invention discloses an ophthalmic disease diagnostic system based on optical coherence tomography (OCT), comprising: an image data acquisition module for acquiring OCT image data; a feature extraction and similarity calculation module for extracting OCT image feature representations and layered structural features of the macular region, calculating macular structural similarity information between different OCT image data, and constructing a structural similarity alignment matrix reflecting the structural similarity among all OCT images; simultaneously constructing an association matrix representing the correspondence between OCT images and disease categories; and a model training module that introduces the structural similarity alignment matrix and association matrix as joint supervision signals into the objective function, adaptively learning to obtain an ophthalmic disease diagnostic model; the ophthalmic disease diagnostic model outputs a classification diagnostic result representing the probability that the input OCT image data belongs to various ophthalmic diseases. This invention achieves automatic identification and classification of ophthalmic diseases, with high diagnostic accuracy and good robustness.

The application discloses a kind of multi-dimensional data intelligent matching method and system, including the following steps: collection multi-source heterogeneous data set, and execute pre-processing;Perform feature extraction, and calculate the correlation strength between features, generate feature weight matrix;Calculate structural similarity, semantic similarity and statistical similarity, and execute weighted fusion according to feature weight matrix;Nonlinear resonance optimization is executed, and stable semantic resonance spectrum is obtained;Establish reinforcement learning matching strategy network, and train matching strategy network;Match strategy network is used after training in stable semantic resonance spectrum Dynamic matching decision is executed, and matching result set is generated.The application fuses semantic resonance spectrum and reinforcement learning method, realizes the intelligent matching of multi-dimensional heterogeneous data, with the advantages of high precision, strong self-adaptation and stability.

The invention provides a response duration analysis method and device, a program product and electronic equipment, and relates to the technical field of computers. The method comprises the steps of recording a target video in response to a target application program starting operation; performing video analysis processing on the target video to obtain an image frame sequence; setting a starting reference image and an ending reference image according to the image frame sequence; determining a starting frame image according to the structural similarity value of the frame image in the image frame sequence and the starting reference image; according to the structural similarity value of the frame image in the image frame sequence and the ending reference image, determining an ending frame image; and determining the response duration of the target application program according to the frame number between the start frame image and the end frame image. According to the method, the starting point and the ending point are automatically positioned in the automatically recorded target video by using the image structure information, so that the response duration of the target application program is accurately determined.

This invention provides a system and program that facilitates the process of generating and / or visualizing molecular networks for library molecular structure content. [Solution] In system 500, the molecular network generation system 502 for library molecular structure content includes an evaluation component 512 that performs a comparison between a first molecular fingerprint including first molecular structure data of a first molecular structure and a second molecular fingerprint including second molecular structure data of a second molecular structure, and a visualization component 516 that generates display data that visualizes a representation of the structural similarity score obtained from the first molecular structure, the second molecular structure and the comparison. The representation by the visualization component includes edges corresponding to the structural similarity score that extend between pairs of nodes corresponding to the first molecular structure and the second molecular structure.

The invention relates to the field of tea image reconstruction, in particular to a tea image bit and spatial resolution joint reconstruction method based on frequency domain feature decoupling. According to the scheme, the method comprises the steps that an original low-bit low-spatial-resolution tea image is acquired and preprocessed, a tea image reconstruction neural network is constructed, and the tea image reconstruction neural network comprises a shared weight module, a non-shared weight module and a frequency domain fusion module; training a tea image reconstruction neural network by using the preprocessed low-bit low-spatial-resolution tea image; after training is completed, all parameters of the shared weight module, the non-shared weight module and the frequency domain fusion module are frozen, a to-be-reconstructed low-spatial-resolution low-bit tea image is input into the trained neural network, and a final reconstructed image is output; and evaluating the reconstruction quality of the tea image through the peak signal-to-noise ratio and the structural similarity index, and comprehensively evaluating the reconstruction effect of the tea reconstruction neural network in combination with the subjective visual effect. The method is suitable for tea image reconstruction.

The invention discloses a monitoring video delay sweep generation method based on scene change self-adaptive frame extraction, and belongs to the technical field of video monitoring and streaming media processing. The method comprises the following steps: acquiring a sweep strategy containing a time period, a reference interval, a difference threshold value, a forced retention period, a window length and a quality threshold value; receiving a real-time code stream or reading a storage side key frame intermediate file; calculating an inter-frame difference degree by using a brightness channel of a multi-scale structure similarity algorithm, determining a frame extraction interval in combination with a static duration, and marking candidate frames; performing quality detection according to definition, brightness, contrast, stripes and noise, and rejecting unqualified products; when a forced period is reached, a bottom frame is reserved; and performing segmented coding according to a time window, and performing assembly output at the end of a time period. According to the method, online and offline processing links are unified, key information is reserved, and redundancy and time delay are reduced.

The invention discloses an accelerator automatic synthesis method, device and equipment, a medium and a product. The method comprises the steps of generating a candidate region set according to a global program structure tree corresponding to a target application program and LLVM intermediate representation; target areas needing to be accelerated are selected from the candidate area set based on preset constraint conditions, and a target area set is formed; and integrating and multiplexing the accelerators of each target area according to the structural similarity of each target area in the target area set to obtain the accelerator finally corresponding to each target area. According to the method, omission caused by manual selection in the prior art is effectively avoided, the overall performance optimization space is improved, the search efficiency is remarkably improved while the optimal performance is ensured, and the hardware resource cost is remarkably reduced.

The present application relates to the technical field of medical image processing, in particular to a low-dose CT image deep learning reconstruction method and system, the method comprising: acquiring paired image samples; constructing a differentiable lesion detectability proxy to derive a detectability gradient from the output image of the generator; performing mapping transformation on the low-dose CT image through the coding-decoding architecture generator, the discriminator and the perception loss function to obtain an enhanced image; using a double-time-scale gradient scheduling strategy to fuse the adversarial gradient, the perception gradient and the detectability gradient to update the generator, and the spatial distribution of the detectability gradient emerges a lesion region of interest attention map during the training process and feeds back to the generator condition input; calculating the signal-to-noise ratio, the structural similarity index and the lesion detectability index of the enhanced image; and outputting a spatial positioning rescan suggestion for those below the diagnostic threshold.

According to the digital twinborn scene illumination dynamic tuning method and device, the electronic equipment and the medium provided by the invention, the digital twinborn scene is rendered through the plurality of illumination parameters of the image rendering engine, and the rendered digital twinborn scene is obtained; acquiring a virtual image in the rendered digital twin scene based on equipment acquisition parameters corresponding to the real image; weighting the pixel loss value and the structural similarity loss value between the real image and the virtual image to obtain a difference value between the virtual image and the real image; and if the difference value exceeds a preset difference threshold value, adjusting each illumination parameter based on the difference value to obtain a plurality of adjusted illumination parameters. The illumination parameters are adjusted through the difference value between the real image and the virtual image, automatic adjustment of the illumination parameters is achieved, it is ensured that the illumination effect of the digital twinborn scene is consistent with the illumination effect of the real scene, and the illumination parameter adjustment efficiency and precision of the digital twinborn scene are effectively improved.