537 results about "Image domain" patented technology

The invention relates to a feature enhancement-based multi-modal image-text tampering detection and positioning method, which comprises the following steps of: obtaining a tampered image-text pair, respectively preprocessing an input image and an input text, and constructing a data set; constructing a multi-modal image-text tampering model, wherein the multi-modal image-text tampering model comprises a text encoder, an image encoder, an image frequency domain enhancement module, a text enhancement module, a multi-cross attention fusion module and a multi-task tampering detection and positioning head; the image frequency domain enhancement module is used for selecting an optimal frequency band and sending the optimal frequency band out of the module; performing feature fusion on the frequency domain feature output by the image frequency domain enhancement module and the image domain feature output by the image encoder to obtain an image enhancement feature; fusing the aligned image enhancement features and text enhancement features by using a multi-cross attention fusion module; and training a multi-modal image-text tampering model by using the data set, and detecting and positioning multi-modal image-text tampering. According to the invention, the feature extraction capability of the network is enhanced, and the tampering detection and positioning capability is improved.

The invention discloses a deep forgery detection method based on a visual language model, and relates to the field of image forensics. The deep forgery detection method based on the visual language model aims to combine multi-source information to improve the discrimination capability of the model on a real image and a generated image. The method comprises the following steps: firstly, extracting image features through an image encoder of a pre-trained CLIP model; meanwhile, a frequency domain enhanced counterfeit perception adapter is embedded in the image encoder to mine potential anomalies of counterfeit images in the image domain and the frequency domain. Secondly, a manual feature extraction module is provided, discriminative low-dimensional features are extracted from the four aspects of the edge, the texture, the frequency and the symmetry of the image, and the discriminative low-dimensional features are used as auxiliary information input in the forgery detection process, so that the robustness and the interpretability of the model are improved; meanwhile, the text cue words are converted into feature vectors through a text encoder of a pre-training CLIP model; and finally, the model predicts a forgery score by calculating the cosine similarity between the image features and the text features so as to realize the discrimination of the authenticity of the image. According to the method, the problem that the detection capability of the model on the cross-dataset is insufficient is effectively improved.

The invention relates to an underwater image enhancement method based on wavelet Mama, which aims at the problems of color shift, low contrast ratio and fuzzy details of an underwater image, extracts shallow layer features of an underwater low-quality image, inputs the shallow layer features into a multi-scale coding network, and performs joint enhancement on low-frequency color information and high-frequency detail features by using a wavelet Mama unit. Global semantic features are obtained by combining down-sampling layer-by-layer compression, and global modeling of color offset correction and contrast enhancement is completed; the spatial resolution is recovered through up-sampling, and reconstruction and enhancement of texture details and color information are realized in combination with jump connection and a wavelet Mama unit; and finally, the features are mapped to an image domain and fused with input image residual errors, and an enhanced underwater image with natural color, clear texture and balanced contrast is generated. According to the method, frequency domain-space domain joint feature modeling is realized through the wavelet Mama unit, and the color authenticity, the structural definition and the detail perceptibility of the underwater image are remarkably improved.

The invention relates to the technical field of constructional engineering, and discloses a concrete crack intelligent identification and analysis platform based on image and point cloud fusion, the platform operates a concrete crack intelligent identification and analysis method, and the method comprises the following steps: S1, synchronously collecting image data and point cloud data of a concrete structure in the same scene; s2, establishing a unified world coordinate system and generating a depth map corresponding to the image; s3, generating image domain crack candidates; s4, generating a depth domain crack candidate; s5, performing weighted fusion on the image domain crack candidate and the depth domain crack candidate to generate a fusion crack response; s6, extracting a crack skeleton; s7, obtaining a crack three-dimensional model; and S8, selecting an optimal view angle to trigger re-acquisition of the crack area. Through a closed-loop feedback mechanism, an optimal view angle is selected for re-acquisition by calculating a comprehensive utility value after preliminary acquisition, so that information insufficiency caused by illumination, angle or sparse data is effectively made up.

The invention relates to a remote sensing image domain generalization semantic segmentation method based on multi-scale instance decoupling, which relates to the technical field of remote sensing image interpretation and comprises the steps of multi-scale instance coding, domain invariant instance decoupling and domain generalization semantic decoding. Image features in a segmented backbone network are used as input, and a multi-scale instance coding module effectively senses feature responses of cross-domain instances through deep convolution, multi-scale cooperation and multi-path sensing and is not affected by scale differences, spatial deformation and noise; designing a domain invariant instance decoupling module to decouple the cross-domain style from the image representation; and finally, the decoupled features are input into a domain generalization semantic decoding module so as to predict a segmented image on a target domain which is not seen, and the segmentation method can generalize various remote sensing images which are not seen so as to obtain better segmentation capability.

The invention relates to the technical field of computer vision and artificial intelligence, and discloses a computer vision-based hundred-million-pixel panoramic situation awareness method and system, and the method comprises the steps: synchronously collecting original local field-of-view images through multiple cameras; independently executing target detection and semantic segmentation in an original image domain; constructing local-to-global coordinate mapping based on the calibration parameters; projecting the target features to a global coordinate system and detecting a cross-view-field target; extracting multi-view-field texture, edge and color features, and performing weighted fusion according to a visible area; according to the method, through identification preposition and semantic level feature fusion, geometric splicing distortion is avoided, the cross-boundary identification precision is improved, the calculation load is reduced, target continuous tracking and situation real-time output in a high-density scene are realized, and the method has high engineering suitability and industrial popularization value.

Methods and systems for assisting a vehicle to park using mixed-domain image data. Image-domain data is generated based on raw image data received from a plurality of cameras mounted on a vehicle. The raw image data is associated with a parking zone outside the vehicle, and the image-domain data is generated by a feature-detection machine learning model. A bird's-eye-view (BEV) image is generated based on the raw image data, wherein the BEV image is a projected image of the parking zone. BEV-domain data associated with the BEV image is generated. The BEV-domain data includes data associated with parking landmarks in the parking zone. A computing system localizes the vehicle within the parking zone based on the BEV-domain data and the image-domain data to generate localization data. The computing system performs mapping of the parking zone based on the BEV-domain data, the image-domain data, and the localization data.

The invention provides a magnetic resonance image reconstruction method and device, a model training method and device, equipment and a medium. The method comprises the following steps: inputting acquired to-be-reconstructed under-sampling k-space data into a pre-constructed image reconstruction model, and carrying out continuous k-space reconstruction processing for multiple times to obtain a magnetic resonance reconstruction image; wherein the k-space reconstruction processing in the image reconstruction model comprises the steps of performing frequency domain restoration on input k-space data based on a Transformer encoder introduced with symmetric weight, performing image domain signal restoration on the k-space data after frequency domain restoration, and performing data consistency processing on the k-space data after image domain signal restoration. According to the method, the image reconstruction model is constructed in an end-to-end expansion training optimization mode, so that the frequency domain and image domain information of the undersampled data can be fully utilized, the accuracy of image reconstruction is effectively improved, the computing resource overhead and the time cost can be effectively reduced, and the high efficiency of the reconstruction process is ensured.

The present disclosure relates to techniques for efficient development of initial models and efficient model update and / or adaptation to a different image domain using an adaptive learning framework. For efficient development of initial models, a two-step development strategy may be performed as follows: Phase 1: Model preconditioning, where an artificial intelligence system leverages existing annotated datasets and improves learning skills through training of these datasets; and Phase 2: Target-model training, where an artificial intelligence system utilizes the learning skills learned from Phase 1 to extend itself to a different image domain (target domain) with less number of annotations required in the target domain than conventional learning methods. To efficiently perform model update and adaptation to new datasets after initial model development, a digital pathology scenario is identified, an adaptive-learning method is selected based on the scenario, and the model is updated and adapted to new datasets using the adaptive-learning method.

The invention discloses a privacy protection constrained passive unsupervised graph domain adaptive mechanical fault diagnosis method and device, and the method comprises the steps: inputting a mechanical vibration signal into a pre-trained mechanical fault diagnosis model, and outputting a diagnosis result; the mechanical fault diagnosis model is obtained by training a passive domain multi-scale image domain adaptive network by adopting training data, and the passive domain multi-scale image domain adaptive network comprises a multi-scale feature extractor module, a topological graph convolution module and a feature classifier module which are connected in sequence; the training process comprises two stages of performing pre-training by using source domain data and performing unsupervised domain adaptive training by using target domain data, and a loss function of the training process comprises a label smooth cross entropy loss function for the pre-training process. And the unsupervised domain adaptive loss function used for the unsupervised domain adaptive training process comprises neighbor clustering loss, regularization loss and label prediction diversity loss. The objective of the invention is to realize complete representation and generalization of fault features by deeply mining topological structure information, and optimize decision boundaries to significantly improve inter-domain alignment precision in a passive domain adaptive process, thereby solving the problem of cross-working condition diagnosis of mechanical equipment under data privacy protection constraints.

The invention relates to a remote sensing image domain adaptive classification method based on high-order prototype guidance. The method comprises the following steps; constructing a cross-scene remote sensing image classification data set; constructing a high-order category prototype generation network based on feature interaction; constructing a high-order causal prompt generation network based on prompt learning; and training and testing a remote sensing image domain adaptive classification method based on high-order prototype guidance. Compared with the prior art, through the cross-scene remote sensing image classification network, the problem that in an existing remote sensing image classification task, due to the fact that style differences exist between different satellite remote sensing images, a source domain training model is difficult to migrate to a target domain data set is solved. In addition, key causal features are decoupled by adopting a prompt learning method based on high-order prototype guidance, so that an innovative method for effectively migrating a model trained on source domain satellite data to target domain satellite data is realized.

The invention discloses a hyperbolic space open set semantic segmentation method based on image language supervision. The method comprises the following steps: constructing a hyperbolic space training framework of an image domain and a language domain, and training the hyperbolic space training framework; constructing an open scene semantic segmentation model according to the training framework of the image domain and the language domain; and inputting a to-be-segmented image into the trained open scene semantic segmentation model to obtain a semantic segmentation result. According to the method, the characteristic that the hyperbolic space can encode a hierarchical structure is utilized, and an image and text encoder of a large-scale language vision model is directly migrated from an image-level classification level to a pixel-level classification level in a manner of finely adjusting the large-scale language vision model in the hyperbolic space; on the basis that the open scene capability is reserved, the large-scale language vision model has the semantic segmentation capability.

The invention relates to the technical field of image classification, and discloses a hyperspectral image domain generalization classification method, system and device and a medium. The method comprises the following steps: acquiring a plurality of hyperspectral images of different ground feature types; obtaining a mean value and a variance of each hyperspectral image in a channel dimension, performing random disruption, and determining a spectral variation parameter of the hyperspectral image according to the mean value and the variance before and after random disruption so as to generate a corresponding spectral variation image; separating the center and the background of each hyperspectral image to obtain respective center image and background image, randomly disorganizing the center image and the background image, and generating a corresponding spatial variation image by combining the disorganized center image and background image of each hyperspectral image; a hyperspectral image classification model is obtained by training an enhanced hyperspectral image obtained by fusing a plurality of corresponding spectral variation images and spatial variation images, so that ground feature classification is performed on a to-be-classified hyperspectral image, and the domain generalization performance of the hyperspectral image classification model is improved.

The invention provides an unsupervised CT image denoising method and device based on multi-mode large model text prompt. The method comprises the following steps: performing preliminary denoising on an original CT image by using a pre-training coding and decoding model to generate a first denoised CT image; inputting the first de-noised CT image and a preset text into a multi-modal visual language large model to generate quality cues and detail cues about the first de-noised CT image; denoising the first denoised CT image again by using a trained generation diffusion model according to the quality cue words and the detail cue words to generate a second denoised CT image; inputting the second de-noised CT image serving as an initial de-noised CT image into a three-domain consistency iteration framework for fidelity, and obtaining a final de-noised CT image when an iteration stop condition is met; wherein the three domains comprise an image domain, a projection domain and a wavelet domain. The method can be conveniently applied to various noisy CT images by utilizing the emergence capability and the characterization capability of a large model and combining and designing a denoising frame.

The invention belongs to the field of industrial CT image processing, discloses an extremely sparse angle CT reconstruction method of a multi-stage processing neural network based on a double-domain joint mode, and is used for solving the problem that a traditional reconstruction method cannot reconstruct a high-quality CT image at an extremely sparse sampling view angle. The reconstruction method comprises the following steps: firstly, carrying out extremely sparse sampling on full-angle projection data A acquired by a computed tomography system to obtain projection data B; interpolating the projection data B, and inputting the interpolated projection data B into the trained sine domain network to obtain optimized projection data C; secondly, preliminarily reconstructing the projection data C into a CT image D by using a fast iterative soft threshold shrinkage algorithm (FISTA); and finally, optimizing the image D through a trained image domain network to obtain a high-quality CT image E. According to the method, the CT image under extremely sparse angle sampling can be reconstructed, and the reconstructed CT image has higher structural integrity and definition.

The invention discloses a computed tomography reconstruction method and a related device, and relates to the technical field of image reconstruction, and the method comprises the steps: obtaining original projection domain data; de-noising processing is carried out on the original projection domain data to obtain projection domain data after de-noising processing; inputting the denoised projection domain data into a computed tomography reconstruction model to obtain computed tomography of an image domain output by the computed tomography reconstruction model; and enhancing the computed tomography of the image domain output by the computed tomography reconstruction model to obtain the final computed tomography. According to the method and the device, the improvement of the efficiency and the effect of computed tomography reconstruction is realized based on projection domain data denoising, computed tomography reconstruction model and enhancement of a model result.

The present application relates to an image processing method, apparatus, computer device, and storage medium. The method includes: inputting a first image in a first image domain into a to-be-trained image processing model to obtain a second image in a second image domain; acquiring a first image feature corresponding to the first image and a second image feature corresponding to the second image; obtaining a first model loss value based on a target feature correlation degree between the first image feature and the second image feature; acquiring a first eigenvalue distribution corresponding to the first image feature and a second eigenvalue distribution corresponding to the second image feature; calculating a distribution correlation degree between the first eigenvalue distribution and the second eigenvalue distribution, and obtaining a second model loss value based on the distribution correlation degree, where the second model loss value has a negative correlation relationship with the distribution correlation degree. Using this method can improve the image processing effect. The image processing model in the present application can be an artificial intelligence-based neural network model, and an artificial intelligence cloud service is provided based on this model.

The invention discloses an emotion feature fused bimodal forged information detection method, which comprises the following steps of: extracting emotion features, respectively extracting emotion categories of a text and an image through a large language model, respectively giving emotion intensity of the categories, and scoring the emotion; analyzing the emotion consistency of the mixed modal features of the text and the image; the hybrid expert network calculates domain probabilities to which a text mode, an image mode and a hybrid mode belong respectively through a text domain encoder, an image domain encoder and an image text hybrid feature domain encoder based on domain class processing features; the hybrid expert network learns feature information of a text mode, an image mode and a hybrid mode according to the domain information; the AdaIN network adjusts modal distribution, the AdaIN network adjusts feature distribution according to contribution of each modal, the adjusted features are fused and input into the classifier, a final classification result is obtained, and the information detection accuracy is greatly improved.

The invention relates to a cultural relic virtual restoration method, and specifically, the method comprises the steps: firstly, depicting the global correlation of a cultural relic image through low-rank Tucker decomposition and Stiefel manifold; then, Gaussian noise is removed by utilizing a Frobenius norm, and meanwhile, sparse noise is removed by adopting an L0 norm aiming at the deficiency of an L1 norm; furthermore, a weighted L0 norm is introduced into a gradient image domain to explore a local segmentation smooth structure of the cultural relic image. In addition, the patent designs an ADMM algorithm based on a hard threshold to solve the provided model. In each step of algorithm iteration, a rank increase strategy is adopted to dynamically adjust a tensor rank, so that the burden of searching an optimal tensor rank is avoided, and the calculation cost is reduced. Therefore, the damaged area of the cultural relic is complemented and repaired, and the original appearance of the cultural relic is recovered while the cultural relic is protected.

The invention provides a spectrum self-attention and cloud mask guided remote sensing image cross-modal cloud detection method. According to the method, the spectral self-attention sub-network in the remote sensing image cross-modal cloud detection network is constructed, and the dependency relationship between different spectral bands of the remote sensing image is effectively mined and utilized, so that the recognition accuracy of cloud region features is improved. According to the method, cross-modal migration from the source domain image to the target domain image is realized by using the migration sub-network of the unsupervised image domain in the remote sensing image cross-modal cloud detection network, and the multi-spectral image with the target domain features is generated, so that the defect of the source domain image in spectral information is overcome, the false detection and omission ratio is effectively reduced, and the detection efficiency is improved. And the robustness of cloud detection of the model in a complex scene is enhanced.

The invention discloses a sparse view angle 3D-DSA reconstruction method based on a three-dimensional Poisson generative model. The method comprises the following steps: obtaining pairing data of a sparse view angle 2D-DSA projection drawing and a 3D-DSA reconstruction image; extracting features of a projection image by using a projection domain encoder, then converting features of two-dimensional projection into a three-dimensional image domain according to a geometrical relationship of the cone beam CT, and then obtaining a prior image by using an image decoder; the method comprises the following steps: constructing a three-dimensional Poisson generation model, adding noise to a three-dimensional image patch in a training stage to obtain a disturbance image, outputting a network reconstruction image by taking a prior image as a condition and the three-dimensional image patch before noise addition as a target image, and calculating loss of the output image and the target image to update network parameters; the mean square error loss and the mean square error loss of the maximum intensity projection images of the three orthogonal planes are used during loss calculation; in the sampling stage, random noise is used as input, a prior image is used as a condition, noise of a noise image is continuously denoised within a limited step length, and finally a reconstructed 3D-DSA image is obtained.

The invention discloses an unmanned aerial vehicle rescue full-coverage search and rescue method and system, and the method comprises the steps: enabling a search and rescue boundary to deviate inwards for a certain distance, obtaining a search and rescue region boundary deviation line, judging whether the center of a region defined by the search and rescue region boundary deviation line is provided with a hole or not, and if yes, carrying out the search and rescue. When no hole exists, the area is mapped to a circular ring containing a slit through conformal slit mapping by adopting annular slit mapping, the boundary of the area is mapped to the boundary of the inner side and the outer side of the circular ring and the inner arc-shaped slit, and when no hole exists, the area is mapped to a disc containing a slit through slit mapping by adopting disc-shaped slit mapping, and the circular ring is formed. The boundary of the area is mapped into the boundary of the outer side of the disc and an internal arc-shaped slit; unequally-divided spiral lines which are not intersected with the slit in the circular ring or disc image domain are obtained; the unequally divided spiral lines are mapped back to an original image domain through inverse mapping of conformal slit mapping to obtain a spiral track on an original image, search and rescue are carried out along the boundary offset line of the search and rescue area, and dangerous areas which are not searched and rescued through the spiral track are checked.

The invention provides a magnetic resonance image reconstruction method and device, an imaging system and a storage medium, and the method comprises the steps: inputting to-be-reconstructed undersampling K space data into an image reconstruction model, and obtaining a reconstructed image; and the image reconstruction model is trained by taking model training loss value minimization as a target: performing inverse Fourier transform and down-sampling on the full-sampling K-space data to obtain a full-sampling image and under-sampling K-space data, and performing randomization, interpolation and inverse Fourier transform on the under-sampling K-space data in sequence to obtain a frequency domain training image; and fusing an undersampling training image obtained by inverse Fourier transform of the undersampling K space data and an undersampling image obtained by sequentially carrying out multiple times of second downsampling and interpolation on the undersampling training image to obtain an image domain training image, and calculating a loss value of the training image obtained by fusing the frequency domain and the image domain and the full-sampling image. According to the invention, through adaptive filling and reconstruction of the image domain and the frequency domain, efficient, rapid and high-quality magnetic resonance image recovery is realized.

The invention provides a PCB surface defect detection method and system based on double-layer SAM model collaboration. The method comprises the steps that industrial image domain features and image block features of a to-be-detected PCB image are acquired; performing global semantic segmentation on the PCB image to be detected by adopting an upstream SAM model, extracting a plurality of semantic masks, and representing the semantic masks as weighted semantic mask features; on the basis of the image block features and the weighted semantic mask features, calculating the Euclidean distance between the to-be-detected PCB image and a core memory bank constructed on the basis of the normal PCB image, and obtaining a multi-level anomaly score; fusing the multi-level anomaly scores to obtain a rough anomaly score graph; a downstream SAM model is adopted, the rough anomaly score graph is used as a prompt, and an effective segmentation area mask is obtained in combination with a normal PCB structure learned by the upstream SAM model; and based on the effective segmentation region mask, pixel-level abnormal region refinement is performed on the rough abnormal score map to obtain a fine abnormal score map for category judgment and positioning of defect detection. According to the invention, high-precision defect detection in a small sample scene is realized.

The invention relates to the technical field of medical image processing, in particular to a CT radiography intelligent imaging method and system based on DCT-GAN multi-scale fusion, and the method comprises the steps: obtaining a preprocessed target NCCT-CTA image data set; dividing according to a preset proportion based on the target NCCT-CTA image data set to obtain a training set and a verification set; the training set and the verification set are input into an initial generative adversarial network based on DCT-GAN multi-scale fusion for model training, in the training process, a virtual CTA image is generated by a generator, an input original CTA image and the virtual CTA image are discriminated in an image domain and a frequency domain subjected to discrete cosine transform processing by a discriminator, and a model is obtained. Determining a network overall loss function based on the similarity difference between the image domain and the frequency domain and the discrimination difference of the discriminator on the two domains; and inputting the preprocessed real-time NCCT image into the trained target generative adversarial network to generate a high-quality and high-resolution virtual CTA image.

The invention discloses a short-term power load prediction method based on image domain feature extraction, and the method comprises the steps: obtaining and preprocessing multi-source time sequence data of a power load, and forming a one-dimensional power load historical information sequence data set; converting a one-dimensional time sequence in the data set from a time domain to an image domain by adopting a GAF algorithm to form a two-dimensional power load historical information image data set; inputting images in the two-dimensional power load historical information image data set into a ConvLSTM (Convolutional Long-Short-term Memory) model to extract daily and weekly period and time sequence characteristics of the power load and carry out short-term power load prediction; meanwhile, an extrusion-excitation (SE) channel attention mechanism is introduced into the ConvLSTM model so as to consider the nonlinear influence of the characteristic variable on the power load change. According to the method, the prior knowledge and the similar daily method are utilized to predict the logic of the power load, the power load prediction precision is improved, and automatic intelligent auxiliary reference is provided for power system dispatching decision making.

The invention discloses an urban impervious surface all-weather extraction method and system based on a stacked generative adversarial network, a storage medium and electronic equipment, and the method comprises the following steps: carrying out the corresponding preprocessing of an optical remote sensing image and an SAR remote sensing image, and dividing the images into a training set and a test set; constructing an impervious surface data set: constructing a generative adversarial network for conversion from an SAR image domain to an optical image domain, and constructing an urban impervious surface extraction model based on optical and SAR image feature fusion; and training and testing the model to obtain a trained training model. According to the method, the simulated optical image is generated from the SAR image by using the generative adversarial network, and then the generated simulated optical image and the SAR image are fused and input into the impervious surface extraction model based on the generative adversarial network for impervious surface extraction, so that the impervious surface extraction precision based on the SAR image is improved; all-weather extraction of the urban impervious surface in the cloudy and rainy tropical / subtropical region is realized, and the application blank of optical remote sensing is filled.

The invention discloses an absolute quantitative correction method and device for medical imaging equipment, and relates to the technical field of nuclear medical imaging equipment correction. When the method is executed, the actual activity concentration of the calibration die body injected with the first known tracer activity serves as the first activity concentration; then, the calibration die body injected with the first known tracer activity is collected in medical imaging equipment, and the activity concentration in a first image domain obtained through calculation serves as second activity concentration; and finally, calculating a correction factor based on the first activity concentration and the second activity concentration, wherein the correction factor is used for correcting absolute quantification of the medical imaging equipment. Therefore, the correction factor is calculated by comparing the activity concentration calculated by the medical imaging equipment with the actual activity concentration of the calibration die body, so that the medical imaging equipment can accurately adjust the measurement result according to the correction factor during subsequent absolute quantitative analysis; the effect of improving the absolute quantification accuracy of the medical imaging equipment is achieved.

Diffusion models are machine learning algorithms that are uniquely trained to generate high-quality data from an input lower-quality data. For example, they can be trained in the image domain, for example, to perform specific image restoration tasks, such as inpainting (e.g. completing an incomplete image), deblurring (e.g. removing blurring from an image), and super-resolution (e.g. increasing a resolution of an image), or they can be trained to perform image rendering tasks, including 2D-to-3D image generation tasks. However, current approaches to training diffusion models only allow the models to be optimized for a specific task such that they will not achieve high-quality results when used for other tasks. The present disclosure provides a diffusion model that uses variational inferencing to approximate a distribution of data, which allows the diffusion model to universally solve different tasks without having to be re-trained specifically for each task.

The invention discloses a medical MRI image denoising method, device, system and equipment, a medium and a product, and relates to the technical field of image enhancement. The method comprises the following steps: firstly, converting a medical MRI image signal into an image domain to obtain three-dimensional image data, then selecting three-dimensional image sub-blocks, searching M three-dimensional image similar sub-blocks which are most similar in a predefined neighborhood, and then carrying out vectorization straightening processing and superposition on the sub-blocks to obtain an original matrix Y; then assuming that an original matrix Y is equal to X + N + S and solving the minimization objective function to obtain a de-noising matrix X, reconstructing according to the de-noising matrix X to obtain new sub-blocks corresponding to the sub-blocks, and replacing the new and old sub-blocks to obtain new three-dimensional image data subjected to image de-noising. Therefore, the problem that the denoised image is blurred when the prior art is applied to MRI denoising can be solved, the quality of the medical MRI image is improved, and subsequent image analysis is facilitated.