404 results about "Pixel image" patented technology

The pattern generation method based on diffusion model fine tuning comprises the following steps: constructing a data set; the generation model comprises an edge detection model ControlNet and a stable diffusion model, the stable diffusion model comprises a text encoder CLIP, a diffusion model U-Net and an image decoder VAE, the CLIP text encoder converts an input natural language into word vector features and inputs the word vector features into a U-Net network in a submerged space, and the diffusion model U-Net and the image decoder VAE are connected with the edge detection model ControlNet and the diffusion model U-Net; the U-Net network is controlled to carry out iterative denoising on pure noise in a low dimension to generate a compressed image, and finally the compressed image is restored to an original pixel image through a decoder VAE, so that text content control image generation is realized; carrying out fine tuning training on the generative model by adopting a LoRA parameter efficient fine tuning technology; and intelligently generating a batik style pattern. The method has the advantages that the texture precision of the generated image can be improved, and the personalized style image can be generated.

The invention provides a processing method and system for reconstructing an LDI sub-pixel image through sparse compression and application, and the method comprises the steps: calculating a low-resolution to-be-exposed image corresponding to each phase structure in advance through a computer image compression algorithm; the low-resolution to-be-exposed images obtained through calculation are loaded to the digital micro-reflector according to the time sequence, and the display moment of each pattern is matched with the rotation position of the corresponding phase structure; coupling the pattern of the digital micro-mirror and the phase encoding mask in a frequency domain through a 4-f optical system, and reconstructing a high-resolution exposure pattern at a sub-pixel level by utilizing a diffraction effect; projecting a high-resolution light field of the reconstructed high-resolution exposure pattern to the surface of the photoresist, and forming a target circuit pattern by accumulating exposure dose; and the micro-nano structure with sub-pixel precision is obtained after development. The system comprises an image compression module, a pattern matching module and a pattern forming module. According to the invention, the manufacturing cost of the laser direct imaging equipment is reduced under the condition of the same precision.

The embodiment of the invention provides a multi-modal data fused postoperative plastic surgery infection early warning method and system. The method comprises the following steps: firstly, performing quality evaluation and restoration on postoperative wound images, clinical texts, wearable vital signs and baseline features; performing double-flow coding and priori segmentation on the repaired image to obtain a coarse mask; constructing a super-pixel graph, obtaining a topological consistent fine mask through a graph neural network, and quantifying the erythema / exudation area; the text and the vital signs are coded respectively and then fused with the image features and the masks through cross-attention and self-attention, and a fusion implicit vector and uncertainty are obtained through quality weight gating; an individualized dynamic threshold value is constructed by combining baseline risk, logarithmic probability calibration and a sliding window trend, a confidence band is output by using conformal prediction, and early reliable infection early warning is realized.

The invention discloses a screen shooting traceability method, device and equipment and a readable storage medium, which are applied to the field of computer software, and comprise the following steps: encoding watermark information to obtain a watermark information sequence; embedding the watermark information sequence into the image through an R channel, a G channel and a B channel of RGB (Red, Green, Blue) to obtain an image with a watermark; and acquiring the shot image with the watermark image, and tracing to obtain watermark information based on the shot image. According to the method, the watermark is generated by using the RGB channel of the single pixel image, so that the visual concealment of the watermark image is improved, the difficulty in perceiving of human eye recognition is ensured, and the visual quality of a user is not influenced; the watermark embedding position is not influenced by an original image carrier, and watermark information can be hidden in any area of the image; and the watermark information recovery success rate is high.

The invention discloses a multisource remote sensing image surface water body extraction method based on a graph neural network. The SAR image is used as a main part, and the optical image, the digital elevation model and the local incident angle information are fused. Firstly, radiation correction, geometric correction and noise suppression are carried out on SAR data, and multi-source data spatial resolution matching is carried out. And then converting the image into a graph structure by adopting a self-adaptive Bayesian superpixel segmentation method. Then, geometric, texture and physical features are constructed for the superpixel nodes, and feature dimensions are optimized through a feature selection algorithm; and finally, constructing a classification aggregation graph convolutional neural network, wherein the core lies in a self-adaptive microaggregation mechanism of the classification aggregation graph convolutional neural network. According to the mechanism, neighborhood superpixels can be dynamically divided according to categories and feature fusion is carried out, and the recognition capability of the model on water body boundaries and regional heterogeneity is remarkably improved. According to the method, multi-source remote sensing data, superpixel graph structure conversion and an optimized graph convolutional network are comprehensively utilized, and the precision and robustness of water body extraction in the complex surface environment are effectively improved.

The invention discloses a semiconductor photoelectric chip anomaly detection method and device and computing equipment, and the method comprises the steps: loading a first target chip image, based on a first color model, of a target semiconductor photoelectric chip, and converting the first target chip image into a second target chip image based on a second color model, the second color model comprises a red channel, a saturation channel and a brightness channel; performing superpixel segmentation on the second target chip image to obtain a target chip superpixel image, and collecting a feature vector of each target superpixel in the target chip superpixel image; predicting a plurality of target key areas in the target chip superpixel image by using the target visual large model; based on the feature vector of each target superpixel, generating a corresponding OCSVM model for each target key area; and performing anomaly detection on any target key area based on the OCSVM model corresponding to any target key area. According to the invention, accurate anomaly detection can be carried out on the single key area of the semiconductor photoelectric chip.

The invention discloses a building plane element recognition method based on a graph neural network and a convolutional neural network, and belongs to the technical field of building plane element recognition. The identification method comprises the following steps: inputting a building plane pixel image, and carrying out preprocessing through image layer screening and information extraction to obtain a binary pixel image; vectorizing the binarized pixel image to obtain a vectorized image; performing image segmentation on the vectorized image; using the segmented vectorized image to construct a regional adjacency graph; performing regional adjacency graph optimization on the regional adjacency graph through graph analysis and regional merging by using a graph neural network, and identifying a room edge; calculating the area of the room; using the convolutional neural network image recognition model to recognize articles in the room; and performing room function prediction according to an identification result. According to the method, the problem of shape blurring caused by convolution in a traditional method is avoided. The graph neural network model training effect is improved, and the classification accuracy is improved. The method is suitable for various planar graph styles and is high in universality.

The invention relates to the technical field of image segmentation, in particular to an anorectal focus automatic segmentation method based on deep learning, which comprises the following steps: acquiring an anorectal image pixel map, extracting contrast and direction offset to mark candidate focus points, screening overlapped marks to generate a focus activation mark map, and establishing a response map to generate a boundary response distribution map. And training the network to output a classification graph, and extracting a truncation path to complete image segmentation. According to the invention, through extracting the contrast value and the gradient amplitude of the local gray level co-occurrence matrix, accurate capturing of the spatial difference of the lesion area under a complex background is realized, through constructing a response map and direction consistency comparison mechanism and combining multi-dimensional features such as a direction gradient histogram and a structure tensor, the area discrimination capability and the edge classification precision are improved, and the accuracy of edge classification is improved. The texture stability is judged by means of anisotropic standard deviation, a fuzzy edge mask is set, truncation paths are screened in combination with a main direction vector included angle deviation trend, and continuity and stability of a boundary convergence position are ensured.

The invention relates to the technical field of deep synthesis, and discloses a deep synthesis service instantaneous algorithm, comprising the following steps: preprocessing input data, including data cleaning, format conversion and feature extraction; inputting the preprocessed data into an improved neural network model for feature learning and synthesis calculation, wherein the improved neural network model is optimized by introducing an attention mechanism and a lightweight network structure; a dynamic computing resource allocation mechanism is adopted, and corresponding computing resources are allocated to the deep synthesis tasks according to real-time computing loads and task priorities; and outputting a synthesis result. Through data preprocessing optimization, lightweight model design and dynamic resource allocation, the average delay of the algorithm for processing a frame of 256 * 256 pixel image is reduced to be within 80ms, the requirement of a real-time interaction scene is met, the delay is reduced by more than 60% compared with an existing algorithm, and the real-time performance can be improved.

The invention belongs to the technical field of asphalt paving segregation detection, and particularly relates to an asphalt temperature segregation detection method, system and device. The method comprises the following steps: S1, acquiring an asphalt infrared image; s2, performing super-pixel processing on the asphalt infrared image by adopting a multi-scale morphological watershed algorithm to obtain a super-pixel image; s3, performing clustering processing on the super-pixel image by adopting a K-Means clustering algorithm to obtain a plurality of segmented regions; s4, processing each segmented region by adopting an FCM clustering algorithm to generate a fuzzy label, and merging the fuzzy label into the superpixel image to serve as a segmentation result; and S5, extracting temperatures of different points in the subareas, and judging whether segregation exists in the corresponding subareas or not according to the temperatures. According to the invention, the positioning precision can be improved, so that the asphalt temperature detection requirement in the paving process is met.

The invention relates to the technical field of artificial intelligence and finance, and provides an artificial intelligence-based text wordart method, device, equipment and medium, on one hand, a target text is rasterized based on a text wordart model to obtain a target raster image text, a vector text layer is converted into a pixel layer, and the pixel layer is converted into a pixel layer; the text is not limited by language types, the problem that the Chinese language structure is complex is avoided, and the text-to-graph problem is converted into the graph-to-graph problem; on one hand, a large language model is used for recognizing regional coordinates used for semantization in raster image characters and expanding style cue words, and the controllability degree of text art generation is improved; and on the other hand, the discriminator is utilized to guide the diffusion model to generate a font which can be obviously recognized, and meanwhile, through an adversarial training mechanism, the generated wordart has the advantages of clear font recognition and artistic semantic change, so that the style and the font are highly fused and balanced, and the problems of low efficiency and poor quality of the text wordart are solved.

The invention discloses a power system software vulnerability detection-oriented code multi-modal fusion representation learning method, which comprises the following steps of: firstly, obtaining four modals of a text, a pixel image, an attribute graph and complexity measurement of a power system software code through program analysis; then, modal interaction feature extraction is carried out by using dual-channel convolution and a graph neural network, and information transmission between modals is enhanced; and finally, designing a multi-modal collaborative fusion framework, performing reconstruction coding on the features based on feature mapping and cross-modal reconstruction loss, and fusing the multi-modal feature representation of the power system software code into a unified representation for constructing a vulnerability detection model for the power system software. According to the method, the completeness of modeling data representation of the power system software vulnerability detection model is enhanced through a code multi-modal interactive fusion representation learning method.

The invention provides a method for correcting the layout of a photomask, and is applied to the technical field of semiconductors. According to the method and the device, a plurality of pixel graphs of which the interval lengths do not meet a preset range are screened out based on the interval lengths between adjacent pixel graphs, and then the key sizes of the screened pixel graphs are multiplied by N times according to different proportions, so that a plurality of first test photomasks are constructed; and then determining the optimal magnification ratio of the initial photomask based on the combination of the wafer data, the incremental limit suboptimal solution and the two-side clamping theorem, namely obtaining the target photomask, so that the unexpected effects are that the optimal magnification ratio of each pixel graph in the initial photomask can be rapidly determined only by using a small part of pixel graphs in the initial photomask, and the target photomask is obtained. The manufacturing efficiency and quality of the photomask and the space utilization rate of the photomask and the wafer are improved, and the manufacturing cost of the photomask is reduced.

The invention relates to the technical field of light field image super-resolution reconstruction, in particular to a light field image spatial super-resolution method based on a multi-scale large kernel attention mechanism, which comprises the following steps: inputting a target light field image into a trained spatial super-resolution reconstruction model, and outputting a spatial super-resolution reconstruction image of the target light field image; in the spatial super-resolution reconstruction model, performing feature extraction on an input light field image through a shallow feature extraction network to generate a shallow feature map; respectively extracting angle features, spatial features and epipolar plane image features from the macro-pixel image of the shallow feature map, the sub-aperture image for generating the macro-pixel feature map and the epipolar plane image of the shallow feature map through a large kernel feature extraction network, and fusing to generate a 4D light field feature map; and performing up-sampling reconstruction on the 4D light field feature map through an up-sampling reconstruction network to generate a spatial super-resolution reconstruction image of the input light field image. According to the invention, the reconstruction quality of the spatial super-resolution of the light field image can be improved.

Methods, apparatuses, and software are disclosed for multilayer metrology. One method includes obtaining image data of an object with an SEM system, with the image data acquired at multiple landing energy levels. A composed image is generated by performing pixel-by-pixel image processing of the image data. A metrology characteristic is determined from the composed image and metrology is performed on a feature based on the metrology characteristic.

The invention relates to a laser radar point cloud densification method and device fusing image information, and a medium. The method comprises the following steps: fixing a visual camera and a laser radar on a rigid tool, and carrying out joint calibration and data alignment; a laser radar is used to collect three-dimensional point cloud data, and a visual camera is used to shoot a scene; performing super-pixel segmentation on the image by using an image brightness linear iterative clustering algorithm; mapping the three-dimensional point cloud data into a segmented two-dimensional image superpixel pattern spot region by using a jointly calibrated camera model conversion relationship to realize feature clustering of the original three-dimensional point cloud of the laser radar; performing curved surface fitting on a clustered result by using a random sampling consistency algorithm; and linear interpolation is carried out on the fitted curved surface, dense points which are not covered by the original point cloud of the laser radar are supplemented and generated, and densification of the point cloud of the laser radar is realized. According to the method, the point cloud densification precision and practicability are improved, and technical support is provided for automatic driving, robot navigation and three-dimensional reconstruction.

The invention discloses a blade health degree detection method and system based on transfer learning. The blade health degree detection method comprises the following steps: acquiring real-time vibration data of a fan unit blade; and converting the real-time vibration data into a time sequence oscillogram. And converting the real-time vibration data into a two-dimensional pixel image by adopting a Grubrum angle field algorithm. And identifying the time sequence oscillogram according to a pre-constructed first identification model. And identifying the two-dimensional pixel image according to a pre-constructed second identification model. And fusing the classification result of the first identification model and the classification result of the second identification model based on a weighted voting strategy to obtain a final classification result of the health degree of the fan unit blade. According to the method, high-precision, low-cost and real-time detection of the health degree of the fan blade is achieved, and the defects that in the prior art, only basic structure monitoring is concerned, calculation complexity is high, and dependence on data is high are overcome.

The invention relates to the technical field of video image recognition, in particular to an idle parking space detection method based on maximum matching and parking space segmentation, which comprises the following steps of: S1, calibrating all parking spaces in a picture under the view angle of a camera to form a parking space calibration pixel map and a JSON dictionary configuration file; s2, performing semantic segmentation on an input video stream by using a segmentation model, and performing BIFPNEMA bidirectional cross-scale connection and a weighted feature fusion model by using a picture to obtain a detection result graph of a parking space occupancy condition; and associating the parking space attribute of the calibration pixel map to the corresponding parking space of the detection result map through a mapping relationship, and generating a detection result map with the attribute. And S3, filtering out misrecognized parking spaces and irregularly occupied parking spaces by using a multi-priority matching algorithm on the parking space calibration pixel map obtained by calibration and the detection result map to obtain idle parking spaces and occupied parking spaces. S4, the detected parking spaces are output in a serialized mode, and the attribute information and the state information of the parking spaces are combined and output to a service platform.

The invention discloses a light field image angle super-resolution method based on large kernel convolution attention, and relates to the technical field of image processing, and the method comprises the steps: converting an input sub-aperture image into a macro-pixel image, and extracting the shallow features of a light field image through employing a convolutional neural network; based on the shallow layer features of the light field image, extracting deep layer features of the light field image through a plurality of cascaded multi-scale mixed attention modules; performing angle down-sampling, channel expansion, Pixel Shuffle up-sampling and spatial convolution processing on the deep features of the light field image in sequence to generate a super-resolution macro-pixel image, and converting the super-resolution macro-pixel image into a sub-aperture image. The problems that an existing method is complex in calculation process and low in calculation efficiency are solved.

An image sensor comprising a semiconductor substrate, a plurality of photodiodes, a multilayer reflective stack, and a dielectric layer is disclosed. The plurality of photodiodes is disposed within the semiconductor substrate and includes a first photodiode and a second photodiode adjacent to the first photodiode. The multilayer reflective stack comprises a first material having a first refractive index and a second material having a second refractive index. The dielectric layer has a third refractive index and is disposed between the first photodiode and the multilayer reflective stack. The first material is disposed between the second material and the dielectric layer. The first refractive index is greater than the second refractive index and the third refractive index.

In certain embodiments, an ophthalmic system images an eye region comprising at least one eye. The system includes a camera system and a computer. The camera system includes cameras that yield image portions of the eye region. Each camera is located at a position relative to the eye region and yields an image portion. The computer receives the image portions from the camera system. A first image portion is provided by a first camera, and a second image portion is provided by a second camera. The computer identifies identify target pixels of the first image portion, where the target pixels image a location of the eye region; determines image information of correction pixels of the second image portion, where the correction pixels image the same location of the eye region; and corrects the target pixels using the identified image information.

The invention discloses a deep learning-based DCB substrate internal microcrack water immersion ultrasonic adaptive detection method, which comprises the following steps of: collecting a DCB substrate ultrasonic image generated by water immersion ultrasonic C scanning detection, preprocessing, and constructing a deep learning-based DCB substrate microcrack identification network model; comprising a feature extraction layer, three deep convolution modules introducing an adaptive space attention mechanism, a global average pooling layer and a full connection layer. Training the network model by using DCB substrate ultrasonic detection imaging and classification labels thereof in the training set, adjusting model parameters by using the verification set, and evaluating the performance of the network model by using the test set; and deploying the trained network model into water immersion ultrasonic detection equipment, performing real-time detection on the DCB substrate, generating ultrasonic data, preprocessing the ultrasonic data into a unified pixel image, analyzing the image by using the network model, identifying microcracks and giving out a judgment instruction.

A switchable display system includes a display panel, an image source, a switching signal source, and a processing circuit. The display panel includes pixels containing a first pixel. The image source is used to provide image data. The switching signal source is used to provide a switching signal. The processing circuit is coupled to the display panel, an image source, and a switching signal source, stores a correction algorithm, and is used to drive the display panel to display a three-dimensional or two-dimensional image according to the switching signal. When the processing circuit drives the display panel to display the two-dimensional image according to the switching signal, the processing circuit switches a first brightness of the first pixel to a second brightness, which is determined based on brightness values in the image data of pixels of same color adjacent to the first pixel and weight values of the correction algorithm.

The invention relates to the technical field of control of power electronics and energy storage systems, in particular to a WIPI wireless power dynamic balancing method for grid connection of an energy storage system, which comprises the following steps: constructing a fractional order digital twinborn model, and predicting the charge state and temperature rise of a module; based on the prediction result, a demand matrix and a loss matrix are obtained by using a three-layer GraphSAGE and a twenty-quantum-bit variable component sub-optimizer respectively, and the demand matrix and the loss matrix are fused into a power coupling target according to a preset weight; performing target thresholding to generate a 32 * 32 magnetic-acoustic super-structure pixel map, finishing magnetic field shaping at a microsecond level, and driving a GaN inverter to output a matched phase and frequency at the same time; power flow and temperature data are collected through a high-precision analog-to-digital converter, a power transport matrix is iteratively corrected by adopting an entropy regularization Sinkhorn algorithm, and pixels are updated in real time; and finally, writing the feedback Hash into the alliance chain storage to complete millisecond-level credible traceability. According to the invention, rapid, accurate and traceable balance of wireless energy of the energy storage module is realized.

The invention belongs to the technical field of image processing, and particularly relates to a pet behavior analysis method based on image recognition, and the method comprises the steps: 1, obtaining a video frame sequence, carrying out the superpixel segmentation of each frame of image in the video frame sequence, and obtaining a superpixel image composed of a plurality of superpixel blocks; step 2, constructing a Poisson multi-Bernoulli hybrid tracking link by taking the region association graph as input, and segmenting a trajectory entity set into a time sequence fragment set; 3, constructing a skeleton * object * scene hypergraph by taking the time sequence fragment set as input; and 4, finally outputting a behavior analysis result carrying an object participation type and a scene type according to a feeding action segment and a water drinking action segment in the strong recognition initial result of the interactive behavior. According to the invention, unified modeling and interpretable behavior identification of pets, objects and scenes are realized, and the method has the advantages of high robustness, high precision and real-time performance.

The subject disclosure provides for systems and methods for generating a dense depth map. The method may include determining a first region of interest via input from a first sensor device configured to capture supplemental pixel images. The method may include determining a second region of interest via input from a second sensor device configured to capture supplemental pixel images. The method may include receiving depth measurements from a third sensor device configured to capture depth measurements. The method may include determining a plurality of depth maps from the pixel images from the first sensor device, the supplemental pixel images from the second sensor device, and the depth measurements from the third sensor device. The method may include generating from the plurality of depth maps a foveated region comprising a high-resolution image and a low-resolution image oriented on the periphery of the foveated region.

The invention discloses a dinner plate recognition method and device, equipment, a storage medium and a computer program product, and relates to the technical field of image recognition, and the method comprises the steps: obtaining an initial dinner plate image, carrying out the instance segmentation of the initial dinner plate image through an instance segmentation model, and obtaining a mask image of a target dinner plate in the initial dinner plate image; edge pixel acquisition is carried out on the mask image of the target dinner plate, and an edge pixel image of the target dinner plate is generated; performing feature extraction on the edge pixel image of the target dinner plate through a feature extraction model to generate a target feature vector of the target dinner plate; and performing feature retrieval on the target feature vector of the target dinner plate through a vector retrieval library to obtain a dinner plate recognition result of the target dinner plate. The lightweight dinner plate recognition method is constructed through the instance segmentation model, the feature extraction model and the vector retrieval library, the hardware dependence of dinner plate recognition is reduced, and the universality of dinner plate recognition is improved.

The invention provides a nondestructive testing method for the solidification state of a solid-state battery based on multi-mode ultrasound, and the method comprises the steps: carrying out the ultrasonic surface scanning of the solid-state battery through employing an ultrasonic testing device, and collecting a transmission signal and an echo signal of an ultrasonic wave; calculating the transmission signal sound velocity and the transmission signal peak-to-peak value of each point on the scanning surface of the solid-state battery; mapping to generate an HSV (hue, saturation and value) color pixel map; judging the curing state of the solid-state battery according to the transmission signal sound velocity and the HSV color pixel map; when the curing state is not good, positioning a suspicious defect area according to the HSV color pixel map; the curing defect type is determined from the display of the HSV color pixel map and a B-scan map generated from the echo signal. According to the method, nondestructive detection of the solid-state battery curing state can be realized, and cooperative improvement of detection efficiency and accuracy can be realized.

The invention discloses a wetland human interference activity rapid identification method based on remote sensing and deep learning, and relates to the technical field of image identification, and the method comprises the steps: 1, carrying out the radiometric calibration, geometric registration and pixel-level alignment of a multispectral remote sensing image of a target wetland obtained at different time phases, and dividing the multispectral remote sensing image into image tile units according to a preset size; 2, two image tile units of different time phases at the same geographic position are paired to form a tile pair, and after fusion of a spectral coding branch and a time coding branch, a change coding graph is output and used for representing wetland surface change characteristics; step 3, constructing a double-layer adjacent pixel graph by taking the change coding graph as input, and performing segmentation under multiple scales through a progressive form graph segmentation algorithm to obtain communicated change clusters; and 4, comparing the attribute values of the change clusters with a template threshold value, and carrying out type interference activity type judgment on each change cluster. The method has rapid processing capability and good adaptability.

An image processing system includes an image sensor including a pixel array and a digital signal processor. Pixel signals, are spaced apart from an adjacent signal in a first or second direction by a first interval. The pixel array includes a first pixel in a (k)th row and a (j)th column, a second pixel in a (k+2)th row and the (j)th column, a third pixel in the (k+2)th row and a (j+2)th column, and a fourth pixel in a (k+4)th row and the (j+2)th column, The image sensor outputs first to fourth binning signals respectively by performing binning based on a first to a fourth pixel signals output from the first to fourth pixels. The digital signal processor interpolates an image signal in accordance with a mode and outputs interpolation signals, which are spaced apart from an adjacent signal in the first or second direction by a second interval.