Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

136 results about "Visual structure" patented technology

Visual Structures is the second course in the foundations sequence and the course that refines the understanding of the visual structures vocabulary. The emphasis of the course is on the understanding and application of visual structures as a foundation for future efforts in art and design.

Page structure optimization method and system for PowerPoint

The invention provides a page structure optimization method and system for a presentation file. The method comprises the following steps: identifying visual elements in a presentation page, and carrying out logical relationship analysis on the visual elements to obtain logical relationship information of the presentation page; mapping the visual structure information into a page three-dimensional implicit field to obtain three-dimensional voxels corresponding to the visual elements; calculating a visual focus thermodynamic diagram in the page of the presentation file, and distributing a three-dimensional voxel corresponding to the target visual element to a visual sensitive area; on the basis of the position relation of the visual elements and the first typesetting information, combining an information density energy function to predict a global optimal layout, and optimizing a narrative path in the presentation file according to logic relation information and user narrative preference to obtain third typesetting information, so that the spatial arrangement relation of the visual elements conforms to the narrative logic of the user preference; and executing secondary typesetting on the PowerPoint page based on the third typesetting information to optimize the page structure of the PowerPoint page and improve the page typesetting efficiency.
Owner:珠海必优科技有限公司

Cross-modal interaction image restoration method fusing text semantic guidance and visual structure prior

The invention discloses a cross-modal interactive image restoration method fusing text semantic guidance and visual structure priori, which comprises the following steps of: firstly, acquiring natural language description input by a user and an image to be restored, and generating a semantic segmentation map of the image through a semantic segmentation model; encoding the text and image semantics by using a pre-trained cross-modal encoding model to obtain text and semantic features; guiding a semantic alignment attention module through Prompt to realize deep fusion of multi-modal semantic features and image space features; structural enhancement and regulation of image features are realized by constructing a text guide weight graph, performing element-level modulation on the text guide weight graph and the optimized semantic segmentation graph, constructing a cross-modal structure semantic feature graph and generating a structural modulation factor; a four-stage image restoration network is adopted, and a high-quality restoration image conforming to semantic guidance and structure prior is generated step by step. According to the method, the semantic consistency, the structural integrity and the visual reality sense of an image restoration result are improved.
Owner:NANJING UNIV OF POSTS & TELECOMM

Endoscopic polyp segmentation method based on cross-branch feature enhancement and uncertainty guidance

The invention discloses an endoscopic polyp segmentation method based on cross-branch feature enhancement and uncertainty guidance. The method comprises the steps of obtaining a data set and performing preprocessing; carrying out model design on the cross-branch feature enhancement module and the uncertainty-guided prediction regularization module on the basis of a visual Transform structure segmentation cutting model, and fusing a trainable convolutional neural network encoder branch to obtain an endoscope polyp segmentation model; training the obtained endoscopic polyp segmentation model based on a cross entropy loss function; after a polyp image of a test set is segmented according to the trained endoscopic polyp segmentation model, a Sigmoid function is used for extracting a segmentation result, the performance of the model is verified, and a final endoscopic polyp segmentation model is obtained. The method can enhance the features in a cross-branch manner, thereby improving the diversity of the extracted polyp image features. Meanwhile, an uncertainty-guided prediction regularization module is introduced, so that the uncertainty in the training process is relieved, and the generalization ability of the model is enhanced.
Owner:ARTIFICIAL INTELLIGENCE RES INST OF HEFEI COMPREHENSIVE NAT SCI CENT (ANHUI ARTIFICIAL INTELLIGENCE LAB)

Subject analysis method and apparatus, computer device and nonvolatile storage medium

A subject analysis method and apparatus, a computer device, and a storage medium. The subject analysis method comprises: acquiring an image to be analyzed corresponding to a target subject and a trained multi-modal dialogue language model; on the basis of a trained visual structure, performing feature extraction on said image to obtain an image feature corresponding to said image; on the basis of a trained attention structure, performing data processing on the image feature to obtain an initial diagnosis result corresponding to the target subject; and on the basis of a trained language structure, performing data processing on the initial diagnosis result to obtain a subject analysis result corresponding to the target subject.
Owner:TSINGHUA UNIVERSITY +1

Seabed landslide simulation experiment device and method

The invention discloses a seabed landslide simulation experiment device and method, and relates to the technical field of marine geological disaster simulation, the seabed landslide simulation experiment device comprises a closed test box, and at least one side of the test box is provided with a visual structure; the seabed simulation assembly is arranged at the bottom of the test box, the seabed simulation assembly comprises a bearing part, the gradient of the bearing part is adjustable, and a sediment layer, a trap layer and a storage layer which are sequentially arranged on the bearing part form a landslide mass; a seabed water pressure simulation assembly; a water flow simulation assembly; a pore water pressure simulation assembly; the measuring assembly comprises a camera module and a sensor module, the camera module obtains image information of the seabed simulation assembly through the visual structure, and the sensor module is used for obtaining the pressure condition in the test box and the pressure condition of the landslide mass. According to the seabed landslide simulation experiment device, a complex seabed environment can be truly simulated, so that the prediction accuracy of an experiment result on an actual seabed landslide condition is improved.
Owner:GUANGZHOU UNIVERSITY

Single-frame real-time three-dimensional imaging method based on deep learning

The invention discloses a single-frame real-time three-dimensional imaging method based on deep learning. The method comprises the following steps: constructing a lightweight phase recovery network model based on a physical model; building a multi-vision structure light stripe projection system; and constructing a training data set, completing training of the lightweight phase recovery network model, and deploying the lightweight phase recovery network model by using TensorRT. Three wrapped phase images are obtained in real time through parallel reasoning of three monochromatic view angles; unwrapping the wrapped phase into an absolute phase by using a multi-view geometric constraint and three-dimensional phase matching algorithm; and generating a three-dimensional point cloud with color texture information through phase-height mapping and a color image. According to the method, the deep learning technology and a traditional fringe projection physical model are combined, rapid phase acquisition can be completed only through a single frame of image, motion interference is overcome, and online real-time high-precision three-dimensional imaging of a complex dynamic scene is achieved.
Owner:NANJING UNIV OF SCI & TECH

Multi-feature 3D (three-dimensional) Gaussian reconstruction method based on laser vision

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

Remote sensing image semantic segmentation method based on visual structure information guidance

The invention discloses a remote sensing image semantic segmentation method based on visual structure information guidance. The remote sensing image semantic segmentation method comprises the following steps: selecting a remote sensing data set and a tree surveying and mapping data set as experimental data sets; based on the experimental data set, selecting a visual basic model, and extracting a visual structure image from the remote sensing image of the experimental data set through the visual basic model; based on the visual structure image, constructing a feature co-embedding module to extract a visual feature image in the visual structure image; based on the visual feature image, constructing a structure heuristic feature fusion module to perform feature fusion and structure prediction on the visual feature image to obtain a processed experimental data set; and constructing a target loss function to predict the processed experimental data set. According to the method, the structural image is extracted through the visual model, the construction module extracts and fuses the features, and the processed data set is predicted, so that the accuracy and robustness of remote sensing image segmentation can be improved.
Owner:NANJING FORESTRY UNIV

Micro-expression recognition method, system and device and computer readable medium

The invention discloses a micro-expression recognition method, system and device and a computer readable medium, and belongs to the technical field of deep learning algorithms, and the method comprises the steps: taking a motion optical flow graph between a start frame and a climax frame in a micro-expression sequence as the input of a network model; the optical flow feature map passes through a picture embedding layer, the picture embedding layer segments the optical flow feature map into image blocks, and each image block comprises motion information of different face units; in the visual Transform structure, attention calculation is carried out on the Patch Token in each layer of encoder through a selection learning module, and the selected Token and the global classification Token are connected and input into the last layer of Transform encoder; through cross entropy loss and contrast loss, classification and identification of micro-expression emotion types are realized. According to the method, the problem of micro-expression fine-grained recognition can be solved, and discriminative micro-expression motion features can be learned in limited samples.
Owner:INSPUR SOFTWARE TECH CO LTD

Multi-modal large model driven inertial platform temperature field generation method

The invention belongs to the technical field of precision measurement and intelligent calculation, and particularly relates to a multi-mode large model driven inertial platform temperature field generation method. Comprising the following steps: (1) data acquisition and preprocessing: acquiring a structure appearance image and a multi-working-condition thermodynamic distribution diagram of an inertial platform, and performing space alignment by using affine transformation or perspective transformation; (2) descriptive text generation: inputting a multi-angle RGB image of the inertial platform, extracting spatial features through a pre-trained visual large model, and generating structured text description to describe the inertial platform and the environment thereof in the image; (3) constructing a diffusion model: inputting an inertial platform image and a descriptive text, respectively extracting visual structure features and semantic description features, and controlling the generation direction of a temperature field by utilizing the characteristic of gradual denoising of the diffusion model and combining text semantic guidance; and (4) model training and verification: selecting an optimizer and a learning rate scheduling strategy, and verifying the rationality of a model concern area.
Owner:ZHONGBEI UNIV

Front-end code evaluation method and system, electronic equipment and program product

The invention provides a front-end code evaluation method and system, electronic equipment and a program product, and the method comprises the steps: obtaining a front-end code, carrying out the compiling and rendering of the front-end code, and obtaining a generated front-end page and a target front-end page; performing code extraction processing on the generated front-end page and the target front-end page, and performing redundant node cleaning and code structure tree generation processing on extracted markup language codes to obtain a generated code tree and a target code tree; performing tree structure and element content error calculation processing on the generated code tree and the target code tree to obtain a content structure error; performing visual structure error calculation processing on the generated front-end page and the target front-end page to obtain a visual structure error; and according to the content structure error and the visual structure error, determining and obtaining an evaluation result for generating the front-end code. The scheme can improve the generation accuracy of the front-end code, and can be widely applied to the technical field of artificial intelligence.
Owner:CHINA TELECOM ARTIFICIAL INTELLIGENCE TECHNOLOGY (BEIJING) CO LTD

A method and system for constructing a synthetic image detection framework

The present invention discloses a method and system for constructing a synthetic image detection framework, and belongs to the field of image recognition technology. The method includes: obtaining a certain number of image pairs as a training set; preprocessing each image in the training set to form a one-dimensional sequence and inputting it into the linear layer of the visual conversion model for encoding to obtain a sequence code; extracting the high-exposure negative, frequency domain low-pass filter image and dark channel image of each image in the training set for encoding to obtain an additional code; adding the original image code obtained after adding it to the sequence code and the position code is input into the transformer coding model for fusion, extracting global features and inputting them into a multi-layer perceptron for training, and obtaining the category to which each image belongs. The present application uses a visual transformer structure to process images, adds additional codes, and obtains global features for classification; in addition, an image-to-image generation method is used to obtain pairs of real images and generated images, which greatly improves the model training speed and reduces the dependence on the amount of data.
Owner:BEIJING ACAD OF ARTIFICIAL INTELLLIGENCE

Intelligent equipment asset management system and method based on visual map

The invention relates to the technical field of equipment asset management, in particular to an intelligent equipment asset management system and method based on a visual atlas, and the method comprises the steps: creating a visual structure comprising a floor father node and an equipment child node through a go.js engine, configuring differentiated visual identifiers for equipment in different states, completing the construction of a dynamic tree graph; establishing connection with an asset database, and obtaining equipment state update in real time; when the change of the equipment state is detected, the visual performance of the corresponding node is automatically updated; responding to a floor node click event, and executing unfolding / folding animation display subordinate equipment; receiving a search condition input by a user; highlighting matched nodes in the tree atlas, and automatically folding uncorrelated branches; when the equipment is selected, generating a time axis visual chart containing a state change time point; and the maintenance record document corresponding to each state change is displayed in an associated manner. And the management efficiency and the decision accuracy are improved.
Owner:INSPUR YUNZHOU (SHANDONG) IND INTERNET CO LTD

Visual structure of documents in question answering

A question-answering system that receive a natural-language question includes a database to provide a basis for that answer and a structured-query generator that constructs a structured query from the question and uses it to obtain an answer to the question from the database.
Owner:PRYON INC

Geometrically assisted visual positioning method and system

A geometric structure aided visual localization method and a system implementing the method are provided. The method includes retrieving a 3D map of initialization location data, the 3D map being constrained by visual structures and geometric structures and modeled by a Gaussian mixture model of a set of Gaussian distributions with mapped landmarks; acquiring a series of real-time image frames by a camera of a mobile system; and for each real-time image frame: extracting local features from the real-time image frame; predicting a camera pose corresponding to the real-time image frame; creating a key frame by tracking the predicted camera pose in the 3D map; identifying temporally visible landmarks with respect to the created key frame; acquiring 2D-3D correspondences between the local features and the temporally visible landmarks; and localizing the mobile system by estimating a state of the real-time image frame based on the 2D-3D correspondences.
Owner:HONG KONG UNIV OF SCI & TECH R & D CORP LTD

Swimwear with frame round hole structure

The utility model relates to a swimming suit technical field, specifically disclose a cover cup swimsuit with photo frame round hole structure, including cover cup cover and sponge cover cup, the surface of cover cup cover is equipped with hole body, cover cup cover covers on sponge cover cup, the front end face of sponge cover cup is outwardly convex through hole body, compared with the traditional collection of this technical scheme, and the process of saving position, the cover cup edition technology of wrinkle, this technical scheme replaces the collection of hole body cooperation cover cup protruding or wrinkle, from the effect, the surface has no trace of saving position or wrinkle, improves the aesthetic property, in addition, improves the three -dimensional penetrability of round hole, breaks through the planarization, the single limitation of traditional process, creates novel visual structure, product appearance has more discernibility and innovativeness, from the process, simplifies the complicated collection of saving position calculation, the operation such as wrinkle setting, reduces technical threshold and production difficulty, improves the edition efficiency and the finished product flatness, realizes multiple breakthroughs on structural rationality, appearance uniqueness, production convenience.
Owner:银亚辉

A method for installing a bevel gear pair and a cone box

The present invention discloses a bevel gear pair, which includes an active bevel gear and a driven bevel gear, wherein the active bevel gear has a first marking surface, and the driven bevel gear has a second marking surface, and the first marking surface and the second marking surface are flush when the active bevel gear and the driven bevel gear are meshed and the cone centers coincide. Based on the above structure, the marking surface is equivalent to the visual structure of the cone center of the bevel gear, so that when installing the bevel gear pair, it is only necessary to ensure that the first marking surface and the second marking surface are flush, so that it can be determined that the cone centers of the active bevel gear and the driven bevel gear coincide, and then it can be determined that the bevel gear pair has reached the optimal meshing state, which is not only easy to operate, low in difficulty, high in efficiency, and less in time and energy, but also has high installation accuracy, and can effectively improve the performance of the cone box or other applied products. The embodiment of the present invention also provides a method for installing a cone box, which can significantly improve the installation accuracy and efficiency of the cone box by using the above-mentioned bevel gear pair.
Owner:HANWEI GUANGYUAN (GUANGZHOU) INTELLIGENT EQUIP CO LTD

Transformation method of traditional filter press

The invention discloses a transformation method of a traditional filter press, and relates to the technical field of solid-liquid separation equipment.The transformation method comprises the steps that a rack body of an original filter press is reserved; a pressing plate driving structure of the original filter press is reserved; the original pressing plate is reserved; a pulling plate driving structure and a pulling plate trolley in a pulling plate mechanism of an original filter press are reserved, and a visual structure and a distance measuring structure are arranged on the pulling plate trolley; a filter plate group of the original filter press is replaced by a filter pressing plate structure; a filter cloth monitoring structure is additionally arranged on a rack body of an original filter press; a sludge falling structure is additionally arranged on a rack body of an original filter press; the visual structure is used for collecting an image of the filter cloth assembly, the distance measuring structure is used for monitoring the position of the filter pressing plate, the filter cloth monitoring structure is used for monitoring the position of the filter cloth assembly, and the mud falling structure is used for separating a filter cake from the filter cloth assembly. The method is implemented on the basis of existing equipment, the structure of original matched equipment is not changed, the transformation cost is greatly reduced, and the method is suitable for intelligent upgrading of stock markets.
Owner:TIANEN (SUZHOU) FLUID TECH CO LTD

A machine vision-based reinforced concrete silo construction safety monitoring system and method

The application discloses a kind of based on machine vision's reinforced concrete silo construction safety monitoring system and method, related to the field of silo construction, solve the problem that existing reinforced concrete silo construction safety monitoring system has poor monitoring effect, including warehouse outside monitoring module: to be in visual structure monitoring period first to fifth visual monitoring plane periodic visual image analysis, and according to the analysis result obtains the structure visual monitoring deviation corresponding to target shallow circular silo, warehouse inside monitoring module: respectively to each warehouse inside monitoring area is damaged part monitoring, according to the monitoring result obtains the damaged area area ratio corresponding to each warehouse inside monitoring area, and according to this analysis obtains the warehouse inside area damage coefficient, monitoring early warning module: according to structure visual monitoring deviation and warehouse inside area damage coefficient, target shallow circular silo is carried out construction safety early warning, the application can provide guarantee for the construction safety of reinforced concrete silo.
Owner:CHINA CONSTRUCTION FOURTH DIVISION SOUTH CHINA CONSTRUCTION CO LTD +1

A method, apparatus, device, and storage medium for splitting editable image layers.

This invention provides a method, apparatus, device, and storage medium for editing image layer splitting, relating to the technical field of image processing. The method includes: determining the main subject, text, and background through visual structure analysis; generating an adaptive alpha mask by combining chromatic key features and color protection information to obtain a transparent foreground layer; repairing the background based on the main subject area and text area to obtain a background layer; generating an editable text layer; and finally outputting each layer under a unified canvas and coordinate system, achieving editable splitting of the source image. This invention can improve the integrity of image layer splitting and the visual effect during subsequent recombination.
Owner:SHENZHEN XINBOTONG TECHNOLOGY CO LTD

Anomaly detection in documents with visual cues

Techniques are disclosed for understanding the visual structure and patterns of documents and detecting anomalies in data of the documents based on the understanding of the visual structure and patterns of the documents. In one aspect, a computer-implemented method is provided that includes accessing a set of documents, converting the set of documents to a set of images in a binary format, generating a common feature template based on the set of images, comparing each image from the set of images to the common feature template to identify images with at least one anomalous feature, and outputting the images with at least one anomalous feature.
Owner:ORACLE FINANCIAL SERVICES SOFTWARE

Anti-theft warning device based on visual structure

The utility model provides an anti-theft warning device based on a visual structure, which comprises a sweeping robot, an anti-theft component and a rotating component, the rotating component used for adjusting the observation angle of the anti-theft component is magnetically mounted on the rear side of the upper end of the sweeping robot, and the anti-theft component is mounted on the upper side of the rotating component. Compared with the prior art, the anti-theft camera has the following beneficial effects that by arranging the anti-theft assembly, in actual use, a user can adjust the inclination and the height of the camera according to the observation elevation angle of the camera, after the height of the camera meets the requirement, the user can rotate the camera, and the elevation angle of the camera is gradually changed under the action of the third adjusting joint, so that the anti-theft effect is achieved. By means of the arrangement of the magnetic attraction base, the device can be fixed to a designated position in a magnetic attraction fixing mode for monitoring and theft prevention when needed, by means of the detachable structural design, the practicability of the device can be greatly improved, and the device can be rapidly and freely disassembled and assembled when needed.
Owner:ANHUI SAWA INTELLIGENT TECHNOLOGY CO LTD

Angle steel tower defect identification method based on graph neural network

The invention discloses an angle steel tower defect identification method based on a graph neural network. The method comprises the following steps: step 1, constructing an initial structure graph; 2, constructing visual structure joint features of dominant nodes to form a first node feature set; step 3, constructing label missing perceptual coding features of the hidden nodes to form a second node feature set; 4, performing feature estimation on the second node feature set by using a feature probability migration module to form a third node feature set; 5, correcting the third node feature set to obtain final hidden node features, and forming a fourth node feature set; 6, fusing the first node feature set and the fourth node feature set to generate a graph embedding vector set; step 7, performing defect type prediction on each node; and 8, outputting a defect identification result. According to the invention, the graph neural network and the feature probability migration module are combined to realize structural level defect intelligent identification of the unobservable area of the angle steel tower.
Owner:QINGDAO HAINENG ELECTRICITY EQUIP CO LTD

Information processing program and information processing device

To improve work efficiency at maintenance time of a device to be diagnosed.SOLUTION: When having an operator recognize a component or a region where there is possibly occurrence of abnormality in a device being diagnosed that is obtained by analyzing sound data emanating from the device being diagnosed, the component or the region where there is possibly occurrence of abnormality is displayed by a display unit 24 by superimposing marks 70-73 at corresponding points in visual structure information 61 with which the structure of the device being diagnosed is visually graspable. A list of the components or regions where there is possibly occurrence of abnormality may be displayed by the display unit 24.SELECTED DRAWING: Figure 8
Owner:KONICA MINOLTA INC

Data visualization narrative generation method, device and equipment based on paraphrases and medium

The invention discloses a data visualization narrative generation method, device and equipment based on paramagnetism, and a medium, and relates to the technical field of artificial intelligence. The method comprises the following steps: generating sentence-level parastyle features according to parastyle contents input by a user; generating corresponding plot information for the parastyle features; according to data input by a user, determining data feature information corresponding to the data feature name in the external features; generating visual structure information of features according to the data feature information; generating animation structure information of the parastyle features according to the parastyle features and the plot information, the data feature information and the visual structure information corresponding to the parastyle features; and obtaining to-be-rendered data based on the white features and the plot information, the data feature information, the visualization structure information and the animation structure information corresponding to the white features, and rendering the to-be-rendered data to generate a data visualization narrative picture. The visual narrative content with high personalization is generated according to the generation of the visual narrative driven by the user-defined parastaring.
Owner:ZHEJIANG HERYMED TECH CO LTD

Ignition experiment observation system and use method thereof

The invention discloses an ignition experiment observation system and a use method thereof, and relates to the technical field of fuel combustion experiments.The ignition experiment observation system comprises a shock tube, one end of the shock tube is closed, the other end of the shock tube is open, and a reaction cavity is formed in the shock tube; the visual structures are arranged on the two sides of the shock tube, and the reaction cavity can be observed through the visual structures; the light source module is arranged on one side of the shock tube, and the light source module is used for emitting light into the reaction cavity; the visual module is arranged on one side of the shock tube, the visual module and the light source module are symmetrically arranged, and the visual module is used for obtaining image information in the reaction cavity. By arranging a visual structure on the side wall of the shock tube and introducing an optical schlieren visual camera technology, image data of the ignition process in the shock tube can be completely recorded, so that the position and homogeneity of the ignition in the shock tube can be accurately judged, and required data can be provided for research on heterogeneous ignition in the shock tube.
Owner:SHANGHAI SHIP & SHIPPING RES INST CO LTD +1

Dispensing area acquisition method and device and dispensing method and device

The invention relates to an adhesive dispensing area acquisition method and device and an adhesive dispensing method and device, which are used for installing a functional assembly on a base through retention glue, the retention glue at least comprises a first colloid, and the method comprises the following steps: acquiring edge information of a visual structure of the base; acquiring boundary information of the cavity of the base based on the edge information of the visual structure; obtaining size information of the functional component; obtaining a target placement position of the functional component in the cavity according to the boundary information of the cavity and the size information of the functional component; and based on the boundary information of the cavity and the target placement position of the functional component in the cavity, obtaining a dispensing area of the first colloid. According to the dispensing area obtaining method and device and the dispensing method and device, the dispensing precision and the mounting reliability of the functional assembly are improved, the packaging consistency is ensured, and the product yield and the production stability are further improved.
Owner:CHINAMETAL TECH (HENAN) CO LTD +2

Visual illusion hidden image creation method based on text graph large model

The invention discloses a visual illusion hidden image creation method based on a text graph large model. The method comprises the following steps: 1) extracting an initial feature z0 of a reference image x by using a pre-trained text graph large model; 2) performing inversion on z0 by using a de-noising diffusion implicit model DDIM on the basis of the constructed inversion trajectory to obtain Gaussian noise # imgabs0 # 3), performing sampling on # imgabs1 # by using DDIM on the basis of the constructed reconstruction trajectory to obtain reconstruction characteristics # imgabs2 # 4) constructing a generation trajectory of # imgabs3 #, and performing sampling on a randomly sampled noise signal # imgabs4 # obeying standard Gaussian distribution by using DDIM on the basis of the generation trajectory to obtain hidden space characteristics # imgabs5 #, the length of the generated track is equal to that of the reconstructed track, and each time step of generating the track takes a target prompt text v as a guide condition; 5) decoding # imgabs6 # by using a large model of a text graph to obtain a generated image # imgabs7 # conforming to semantics of v and having an x visual structure
Owner:PEKING UNIV

Suspicious person monitoring method and system for bank security

The invention discloses a suspicious person monitoring method and system for bank security and protection, and relates to the related field of image processing, and the method comprises the steps that a multispectral imaging device collects a person movement image data stream to construct a three-dimensional behavior image data body; constructing a virtual behavior simulation environment, generating a multi-path standard behavior track set, and constructing a behavior inertia reference map; executing cross-time-domain optical flow consistency analysis and local visual structure stability calculation of the data volume to construct behavior trajectory disturbance field distribution, and generating a behavior inertia damage degree distribution matrix based on a difference mapping relation with the atlas; carrying out micro-scale visual enhancement processing, and constructing a micro-action tensity tensor by utilizing sub-pixel-level jitter evolution detection and skin table reflection dynamic fluctuation analysis; and inputting the tensity tensor and the damage degree distribution matrix into a visual risk time sequence evolution channel to generate a suspicious risk evolution curve. The technical problem of limited identification precision in existing suspicious person monitoring is solved, and the technical effect of improving the identification precision is achieved.
Owner:中苏圆科技集团有限公司

Lightweight binocular vision structure three-dimensional displacement and motion trail measurement method and system

The invention discloses a lightweight binocular vision structure three-dimensional displacement and motion trail measurement method and system. The method comprises the following steps: performing binocular calibration on binocular equipment consisting of a master camera and a slave camera to obtain camera parameters; collecting a motion image of the to-be-detected structure with the single square composite target; processing the image to identify a positioning point locking target area and extract angular point coordinates; judging whether the target exceeds the field of view of the main camera according to the angular point coordinates, if so, switching the data of the slave camera, and fusing the data of the double cameras in the overlapping region; and calculating six-degree-of-freedom displacement based on camera parameters, angular point coordinates and switched data to obtain three-dimensional displacement and a motion track. The system comprises a binocular camera, a calibration module, an image processing module, a view field switching module, a calculation module and a display storage module. According to the method, the anti-interference performance and the arrangement efficiency are improved through the single-target design, the measurement range is expanded through switching of the field of view of the master and slave cameras, the efficiency is improved through lightweight calculation, the high-precision, large-range and dynamic measurement requirements can be met, and the method is suitable for engineering structure monitoring and precise manufacturing scenes.
Owner:GUANGXI UNIV +1