85 results about "Dynamic contrast" patented technology

The invention relates to the technical field of substation equipment fault diagnosis, in particular to a substation equipment thermal fault diagnosis method and system based on an infrared image. The invention discloses a substation equipment thermal fault diagnosis method based on an infrared image. The method comprises the following steps: acquiring an equipment temperature distribution image through an infrared thermal imager; carrying out denoising, contrast enhancement and normalization preprocessing on the image; extracting features such as temperature anomaly, temperature gradient and hot spot areas; a YOLOv13 model is adopted to identify the equipment type; inputting the features into a deep learning model for fault classification; and implementing multi-level alarm according to the classification result confidence. According to the method, temperature gradient analysis and a regional dynamic contrast enhancement technology are creatively fused, so that the fault detection precision and the early warning capability are remarkably improved, and intelligent diagnosis and graded early warning of the thermal fault of the substation equipment are realized.

The invention belongs to the technical field of image processing, and particularly relates to an MBR flat membrane surface physical damage visual identification method and system, and the method comprises the steps: carrying out the low-pass filtering of a microscopic gray image, so as to melt a micropore background; calculating a texture variation index based on the outlier degree of the gray values of the neighborhood pixel points and the neighborhood gray dynamic contrast gain; determining a damage aggregation weight by using a space attenuation accumulation value, and eliminating isolated noise points; the texture variation index and the damage aggregation weight are fused through a self-adaptive gating function, and damage confidence is generated; and finally marking a connected damage region based on statistical threshold binarization and morphological processing. According to the method, dense micropore interference can be effectively inhibited, and the physical damage identification precision is improved.

The invention discloses a multi-modal sentiment analysis model based on hierarchical adaptive cross-modal fusion, belongs to the field of natural language processing, and is used for solving the problems that in the prior art, multi-granularity sentiment feature extraction is insufficient, a cross-modal fusion mechanism is rigid, and the distribution difference between different-source modals is large. The method comprises the following steps: firstly, extracting features of texts, audios and visual modalities from original video data, and coding the features into advanced semantic features; secondly, multi-granularity information is fused through a hierarchical feature extractor to generate enhanced single-mode features; then, a self-adaptive cross-modal fusion network with a text as a core is adopted to realize bidirectional interaction and dynamic weighted fusion between modals; further, a dynamic contrast learning mechanism is introduced to align modal distribution in a unified potential space; and finally, inputting the optimized multi-modal features into a classifier and outputting an emotion analysis result. According to the model, through collaborative optimization of multi-granularity feature extraction, adaptive fusion and comparative learning, the accuracy of sentiment analysis and the robustness of the model are remarkably improved.

The present application relates to the technical field of medical magnetic resonance imaging, and discloses an intelligent dynamic contrast-enhanced magnetic resonance parametric imaging method and apparatus. In the method, a pixel-level vision transformer is integrated at the connection position of an encoder and a decoder of a generator to enhance the learning capability for non-local patterns and time series information in medical images, thereby improving the parameter estimation precision in noisy and variable environments. Gradient penalty is introduced into a loss function to optimize the performance of a discriminator, so that the discriminator can more effectively distinguish between real and generated images. A CycleGAN-like model can fully extract and use spatial features while performing time series analysis, thereby providing a more comprehensive and detailed PK parameter map, and providing powerful technical support for improving the accuracy of diagnosis of diseases such as cervical cancer.

The invention discloses a bladder cancer level evaluation method and device based on radiomics and a storage medium. The bladder cancer level evaluation method and device are used for improving bladder cancer level evaluation accuracy and reliability. Obtaining a multi-parameter MRI image of the bladder cancer, wherein the multi-parameter MRI image comprises a T2 weighted image, a diffusion weighted image and a dynamic contrast enhanced image; performing tumor focus part sketching on the multi-parameter MRI image; respectively carrying out feature extraction on the delineated T2 weighted image, diffusion weighted image and dynamic contrast enhanced image by using a radiomics package to generate a bladder cancer feature set; performing feature difference screening on the bladder cancer feature set according to the clinical diagnosis tag; carrying out dimension reduction processing on the screened bladder cancer feature set; constructing a single-parameter MRI classification model and a multi-parameter joint classification model according to the bladder cancer feature set after dimension reduction; and using the single-parameter MRI classification model and the multi-parameter joint classification model to jointly evaluate the bladder cancer level of the to-be-detected data.

The invention provides a scoliosis X-ray image enhancement processing method, relates to the field of image enhancement, and introduces a bone shadow edge intensity modulation factor, constructs gray scale bone shadow dynamic contrast mapping, and realizes preliminary enhancement of a scoliosis X-ray image; extracting a centrum center point sequence on the centerline of the spine, introducing a parameter curvature, and constructing an abnormal gravity kernel function to enhance significant focusing of a high-curvature region and smooth suppression of a low-curvature region in the scoliosis X-ray image; according to the method, centrum distribution density mapping is established, a multi-scale attention response map is constructed, a multi-scale fusion weight is introduced, the multi-scale attention response map and a curvature guide saliency map are fused, and an enhanced scoliosis X-ray image is obtained, so that centrum edges, abnormal bending and density mutation regions are clearer, and clinical readability and diagnosis precision are improved.

The invention relates to the technical field of medical image recognition, in particular to an Alzheimer's disease MRI diagnosis method based on width neural network learning, and the method comprises the steps: 1, carrying out the adaptive ROI cutting of a brain MRI image; step 2, performing dynamic contrast compensation binaryzation; step 3, morphological optimization; step 4, gray scale returning; 5, density peak value guided K-means + + feature extraction is carried out; and step 6, obtaining a width neural network, inputting the final enhanced feature map into the width neural network for final classification prediction, and realizing diagnosis of the Alzheimer's disease MRI. According to the method, a binary segmentation method of dynamic contrast compensation and multi-scale feature enhancement of K-means + + driving are combined, the method is devoted to the hippocampus segmentation problem in image recognition for the Alzheimer's disease and the problem that subtle variation of an early-stage AD patient cannot be captured, and convenience is provided for doctors.

The invention relates to the technical field of liquid crystal display screens, in particular to a liquid crystal display screen dynamic contrast optimization method and system. The method comprises the following steps: dividing a backlight module of a liquid crystal display screen into a plurality of backlight partitions arranged in a matrix, dividing a current frame display image into corresponding image sub-blocks according to the coverage area of the backlight partitions, and extracting the brightness statistical characteristics of each image sub-block to obtain an original brightness distribution matrix; driving the backlight module to perform primary display according to the original brightness distribution matrix, and generating a voltage correction value covering the full quantization bit width of the source driver; and performing cooperative adjustment of reducing the PWM duty ratio and improving the liquid crystal light transmittance on the next frame of to-be-displayed image based on the voltage correction value so as to complete dynamic contrast optimization of the display screen. According to the invention, accurate contrast optimization of the liquid crystal display screen under a dynamic working condition is realized through a cooperative adjustment mechanism of backlight and liquid crystal light transmittance, so that an optimized display picture subjected to double dynamic compensation is output.

The application discloses a double memory dynamic contrast learning clothes-changing pedestrian re-identification method. First, the pedestrian video sequence with an identity label is acquired, and a three-dimensional space-time feature map is extracted through a backbone network; then, spatial aggregation features and time sequence difference features are extracted from the feature map; then, the two types of features are projected to identity and time sequence subspaces respectively, and attention weighted aggregation is performed through double memory banks to generate identity enhanced features and time sequence enhanced features; a similarity priority and diversity completion strategy is used to construct a positive and negative sample set, and a weighted contrast loss is calculated to jointly optimize the double memory banks; a sliding average strategy with diversity regularization is used to update the memory prototype; finally, the double path enhanced features are adaptively fused to generate discriminative feature representation, and a pedestrian matching list is returned through similarity sorting. The application focuses on stable identity features independent of clothes through the design of double memory banks and the use of multiple memory prototypes combined dynamically to represent pedestrians.

A kind of dynamic contrast optimization system and method based on vehicle-mounted multimode vision belongs to image enhancement technical field, in order to solve the problem that prior art will amplify noise when widening dynamic range, strong denoising then loss texture details, and cause block distortion, picture flicker and multi-frame ghosting, and then reduce vehicle-mounted visual perception accuracy, dynamic contrast optimization method based on vehicle-mounted multimode vision is by real-time synchronous acquisition vehicle-mounted visible light, infrared image, according to illumination, definition and noise amplitude, screening and eliminating inferior data, decomposing effective image into three components of illumination, texture, noise, optimizing illumination dynamic range by partition, according to noise level, hierarchical denoising retains texture details, fusion two components corrects image deviation distortion, completes double image registration fusion and smooth adjustment contrast, finally outputs high-quality optimized image, the technical scheme in the application fully restores the detail level of highlight and dark area, and effectively suppresses various imaging noises.

The application relates to a method for identifying granulomatous mastitis and non-mass-like enhancement breast cancer, which comprises the following steps: reading original data of a peak enhancement period of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) collected by a magnetic resonance device, segmenting and labeling a lesion area to obtain a segmentation mask; pre-processing the DCE-MRI image and generating three-slice samples, including spatial resampling, intensity normalization, lesion centering and cutting, continuous three-slice sampling and mask superposition fusion; dividing the data into a training set and a test set according to the patient level, and adopting stratified sampling to maintain category balance; constructing a double-attention classification model based on DenseNet121, integrating a convolution block attention module (CBAM) and a selective kernel network module (SKNet) by adopting a progressive insertion strategy; training the model and testing by adopting a lesion-level voting strategy to obtain the identification result of the granulomatous mastitis and the non-mass-like enhancement breast cancer. The method solves the problem that the deep learning research on the identification of GM and NMEBC on the DCE-MRI is insufficient, and has high classification precision and good calibration.

The invention relates to the technical field of medical image processing, in particular to a fundus angiography image processing system. The system comprises a static atlas initialization unit, a dynamic state evolution unit, a kinetic parameter fitting unit and a fundus dynamics report generation unit. The system constructs an initial static anatomical atlas by acquiring a baseline frame and extracting a static anatomical structure; the core of the method is to analyze a dynamic contrast sequence frame by frame based on the atlas to obtain a time sequence evolution state atlas; under physiological prior constraints, the optimal physiological dynamic parameters are fitted by comparing model prediction with a time sequence evolution graph; and finally generating a fundus dynamics report containing the functional quantitative index and the focus activity heat map. The system can also take the kinetic parameters as historical priori and call the kinetic parameters in a future fitting task so as to dynamically adjust a constraint range and realize individualized closed-loop self-optimization; according to the method, the limitation of traditional static snapshot analysis is overcome, and accurate modeling can be carried out on complex physiological dynamic processes such as perfusion and permeation.

This invention relates to an MRI-based method for classifying benign and malignant intraductal lesions (IDLs) of the breast, comprising the following steps: Step 1) Extracting a set of clinical features from MRI images: age, ADC value, BI-RADS category, lesion shape, margin clarity, and TIC curve type; Step 2) Extracting radiomics features from the 3D lesion segments of the T1 DCE-MRI sequence; Step 3) Reducing the dimensionality of the radiomics features using the Lasso algorithm; Step 4) Fusing the dimensionality-reduced radiomics features with the clinical features into an input vector; Step 5) Outputting benign / malignant probability values ​​through a logistic regression model; Step 6) Applying stratification rules to BI-RADS 4A lesions: older patients or those with low ADC values ​​are marked as high-risk and biopsy is recommended; younger patients with high ADC values ​​are marked as low-risk and follow-up is recommended. This method solves the problem of current dynamic contrast-enhanced MRI's difficulty in accurately identifying BI-RADS categories, especially the BI-RADS 4A high-risk subgroup, thereby reducing missed diagnoses and over-biopsy. It is of urgent significance for overcoming the bottleneck of preoperative risk stratification in IPLs and guiding individualized clinical decision-making.

The invention discloses a fine-grained image retrieval method based on unsupervised depth metric learning. The method comprises the following steps: inputting a fine-grained image to be retrieved into a pre-trained fine-grained image retrieval model to output a retrieval result; the model comprises a sampling module, a global similarity calculation module, a distance matrix calculation module and a block-level feature learning module, and the sampling module adopts a balance sampling strategy to generate balanced small-batch samples; the global similarity module calculates and extracts sample features and then combines cosine similarity and manifold similarity to construct a global similarity matrix; the distance calculation module generates an Euclidean distance between samples; the block-level feature learning module extracts local features through overlapped blocks and a center anchor point, and dynamic contrast clustering is carried out; and finally, combining manifold similarity measurement loss and block-level clustering loss to construct a total loss function, and iteratively optimizing model parameters until convergence. According to the method, fine-grained feature learning under an unsupervised condition is realized, and the retrieval precision is improved through multi-module collaborative optimization.

The application discloses a kind of LED display screen display parameter determination method, device, electronic equipment and storage medium.The method comprises: determining the brightness encoding of image to be displayed;According to the brightness encoding of image to be displayed, determine the picture proportion of target gray level pixel in image to be displayed;According to the picture proportion of target gray level pixel in image to be displayed and the initial correction coefficient of LED display screen obtained in advance, the display parameter of LED display screen is determined.The technical scheme solves the problem of poor display effect caused by LED display screen brightness correction, can guarantee the uniformity of LED display screen display, and improve the dynamic contrast of display screen.

The application relates to the technical field of biosciences, and is a circRNA-disease correlation prediction model construction method and correlation prediction method based on dynamic contrast sampling linear and nonlinear feature resonance and related devices, which comprises the following steps: obtaining a training set and a test set; training a multi-view contrast learning DCFR-CDA model by using the training set; introducing a loss function during the training; ending the training when the value of the loss function is stable; obtaining a circRNA-disease correlation prediction model; testing the circRNA-disease correlation prediction model by using the test set; optimizing model parameters; and outputting a circRNA-disease correlation prediction model meeting test evaluation requirements. The application jointly applies a linear decomposition method and nonlinear decomposition methods of multi-layer perception decomposition and one-dimensional convolution decomposition to generate multi-view embedding compatible with global and local modes, thereby effectively improving the prediction accuracy of the model.

The present application relates to the technical field of display dynamic contrast optimization, and particularly relates to a display dynamic contrast optimization system and method. The method comprises the following steps: obtaining the luminance distribution state of a to-be-displayed picture in a current display period of a display screen, collecting the outputtable backlight luminance range of the display screen, and determining the luminance distribution reference of the to-be-displayed picture on the display screen; based on the luminance distribution reference, dividing the display picture into multiple display partitions, and combining the luminance change requirements of each partition to calculate the partition luminance change load degree; and according to the partition luminance change load degree, adjusting the contrast and luminance; the present application realizes the balance between picture level enhancement and display stability in the display dynamic contrast adjustment process by performing partition analysis on the display picture based on the luminance distribution reference.

This invention belongs to the field of image processing technology, specifically relating to a method and system for visual recognition of physical damage on the surface of an MBR flat sheet membrane. The method includes: low-pass filtering a microscopic grayscale image to melt the micropore background; calculating a texture variation index based on the outlier degree of grayscale values ​​of neighboring pixels and the dynamic contrast gain of the neighboring grayscale; determining damage clustering weights using spatial attenuation accumulation values ​​to remove isolated noise points; fusing the texture variation index and damage clustering weights through an adaptive gating function to generate a damage confidence score; and finally, marking connected damage regions based on statistical threshold binarization and morphological processing. This invention effectively suppresses interference from dense micropores and improves the accuracy of physical damage recognition.

The invention provides a boiler heating surface early warning method and system based on multi-dimensional dynamic comparison. The method comprises the following steps: constructing a multi-dimensional ledger database of a boiler heating surface; based on the multi-dimensional ledger database, utilizing a pre-trained ledger comparison model to carry out multi-dimensional dynamic comparative analysis to obtain a comparative analysis result; and generating corresponding risk early warning information according to the comparative analysis result, and visually displaying the risk early warning information. According to the boiler heating surface standing book comparison method, through real-time / historical data dynamic comparison, hidden defects such as heating surface pipe wall thickness reduction, over-temperature operation and creep over-limit are found in advance, and missing detection of manual troubleshooting is avoided. Meanwhile, the performance difference and the aging degree of the heating surfaces in different areas are quantified, empirical judgment is replaced, and effective data support is provided for maintenance strategies (such as pipe replacement / soot blowing adjustment).

The application discloses a method, system and computer readable storage medium for improving dynamic contrast. The application periodically acquires a picture displayed by a display, then acquires a brightness value of a pixel of the current display picture, calculates a brightness percentage of the current display picture, and adjusts the brightness value of the display according to the brightness percentage of the current display picture. Specifically, when the brightness percentage is 0%, the target brightness of the display is 0%; when the brightness percentage is less than N%, the target brightness of the display is N%, wherein N is an integer closest to the value of the brightness percentage, and the range of N is 1-100; and when the brightness percentage is 100%, the target brightness of the display is 100%.

Embodiments of the present application relate to display technology, and provide a projection display device and method. The projection display device comprises: an image processing unit, configured to determine the current scenario, and if the current scenario is a first scenario, generate, by means of a phase modulation channel in the image processing unit on the basis of an image signal, a first signal and a second signal that are synchronized with each other, wherein the first signal is used for representing the image content of an image frame of the image signal, and the second signal is used for representing the brightness of at least one partition in the image frame of the image signal; a display driving control unit, configured to process the first signal to obtain a first amplitude modulation driving signal; a display image phase execution unit, configured to process the second signal to obtain a phase modulation driving signal; and an imaging display unit, configured to perform imaging display on the basis of the first amplitude modulation driving signal and the phase modulation driving signal. The device provided by the present application improves the dynamic contrast of the projection display device.

The present application relates to the technical field of image processing, and more particularly, to an image transmission optimization method, comprising: acquiring a video image and performing a grayscale processing thereon; and performing a transmission optimization processing on the video image based on a TSS block matching algorithm, comprising: dividing the video image into a plurality of initial image blocks, calculating a texture dynamic contrast and an optical flow motion intensity index of each initial image block, taking a product of a mean value of the texture dynamic contrast and a mean value of the optical flow motion intensity index of all the initial image blocks as a weight, and weighting a number of the initial image blocks to obtain a best image block number of the video image. The present application realizes compression of the video image based on the improved TSS block matching algorithm, improves a video image transmission effect of an intelligent monitoring robot dog, avoids picture distortion, loss and the like in a transmission process, and realizes optimization of image transmission.

The invention discloses a Mini-LED backlight display halo perception boundary determination method and system, and aims to solve the problem that a halo boundary in a high dynamic contrast display scene is difficult to objectively define. The method comprises the following steps: acquiring brightness distribution of a display area and a surrounding dark area in a dark environment; guiding an observer to determine a subjective halo boundary position under a standard observation condition; mapping the position to objective brightness data to extract a boundary brightness threshold value; and establishing a sensing boundary model based on multiple groups of peak brightness and boundary thresholds, wherein the sensing boundary model is used for stably determining a halation region which can be sensed by human eyes under different brightness, partition and dimming strategies. According to the method, the perception boundary is used for replacing a fixed absolute threshold value, evaluation consistency is remarkably improved, closed-loop association of subjective perception and objective data is achieved, and reproducible and computable standardized criteria are provided for halo evaluation, algorithm optimization and production line quality control.

This invention discloses a laser projection display method and system. The method involves configuring the micromirrors of a phase-modulated light device into a high-low periodic structure. Then, based on optical diffraction characteristics, the actual height of the phase-modulated light device under each stage of driving signals is measured. Next, an ideal height set for the micromirrors is defined, which changes linearly with the progressively increasing driving signals. The actual heights that are within a defined error range from the actual heights are identified. These actual heights are matched with the phase level to establish an approximately linear relationship between the driving signal and the changeable phase level. This allows the driving signal to be sent according to the desired linearized height distribution in subsequent designs, achieving an approximately linear adjustment relationship between the change in the actual micromirror position and the change in the ideal micromirror position, thereby improving the dynamic contrast adjustment effect.

To provide a dynamic visual acuity-enhancing composition.SOLUTION: Provided is a dynamic visual acuity-enhancing composition containing cassis anthocyanin as an active ingredient and improving dynamic visual acuity, particularly dynamic contrast discriminability.SELECTED DRAWING: None

The invention provides a cross-platform compatibility test method and device, equipment, a medium and a program product. The method comprises the steps of obtaining multi-modal operation data of a to-be-tested application in multiple test platform equipment; obtaining a first weight coefficient of the to-be-tested application, and obtaining a comprehensive multi-modal operation value according to the first weight coefficient and the multi-modal operation data; constructing a multi-modal comparison sample group based on comparison learning, calculating a dynamic comparison coefficient according to the multi-modal comparison sample group, and obtaining a test abnormal value according to the dynamic comparison coefficient and the comprehensive multi-modal operation value; obtaining a multi-modal feature generated when the to-be-detected application runs on the target equipment, obtaining a target running value according to the multi-modal feature, and determining a feature difference value according to the target running value and the comprehensive multi-modal running value; obtaining a standard difference threshold value and a standard abnormal value, and generating a dynamic difference threshold value according to the standard difference threshold value and the standard abnormal value; and whether the characteristic difference value exceeds the threshold value is judged, so that the cross-platform compatibility information of the to-be-tested application is obtained according to the judgment result, and the test report is generated, and therefore, the test efficiency and accuracy are improved.

A dynamic contrast-enhanced MRI reconstruction method may include: when preparing to inject a contrast agent and during injection thereof, using a DCE-MRI sequence to scan an imaging target, and using a high temporal resolution scanning parameter value pre-inputted by a user to acquire K-space data of each phase with high temporal resolution; based on the acquired K-space data of each phase with high temporal resolution, reconstructing an image of each phase with high temporal resolution; based on a low temporal resolution reconstruction parameter value pre-inputted by the user, subjecting adjacent K-space data of multiple phases with high temporal resolution to summing and averaging in a complex field, and subjecting K-space data obtained after averaging to image reconstruction, to obtain an image of each phase with low temporal resolution. Advantageously, images with high and low temporal resolution can be obtained simultaneously with just a single DCE-MRI scan.

The subject matter includes systems and methods for a deep learning technique applied to dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) scans. The disclosed aims to reduce the dosage of gadolinium-based contrast agents (GBCAs) while maintaining accurate detection and enhancement of BBB openings. A spatiotemporal network (ST-Net) is introduced, combining spatial and temporal networks, allowing for the extraction of diagnostic quality images with reduced GBCAs dosage.

The application relates to the technical field of LCD liquid crystal display screen dynamic contrast control, and discloses an LCD liquid crystal display screen dynamic contrast control method and system, which acquires the driving current and driving voltage of an LED light-emitting area in real time, determines the internal working temperature based on the driving current and the driving voltage, and further calculates the photoelectric conversion efficiency of the LED light-emitting area. Based on the real-time calculated photoelectric conversion efficiency, the application can dynamically adjust the preset dynamic contrast adjustment rule and generate corresponding backlight driving instructions. The method effectively solves the problem that, in the prior art, the photoelectric conversion efficiency and spectral characteristics of an LED lamp bead drift due to the increase of the internal working temperature of the LCD liquid crystal display screen under a non-standard working environment, and the preset dynamic contrast adjustment rule is continuously dynamically disconnected with the actual light output.

The invention concerns a method for determining a filter for a visual equipment intended to be placed in front of the eye of a wearer, the filter being able to improve motion sensitivity of the wearer, the method including the following steps: i. subjecting the wearer to a plurality of visible light spectra having different levels of light intensity within the range of wavelengths from 400 to 600 nm; ii. determining a value representative of a dynamic contrast sensitivity of the wearer for each of the plurality of light spectra; iii. selecting at least one light spectrum based on the values improving the dynamic contrast sensitivity; and iv. determining the filter based on the at least one selected light spectrum.