82 results about "Visual saliency" patented technology

The invention relates to the technical field of remote sensing image analysis and cross-modal retrieval. The invention discloses a remote sensing image text retrieval method based on a remote sensing multi-modal basic model, which applies the large-scale pre-training capability of a CLIP large model to semantic alignment of a remote sensing image and a text by finely adjusting the CLIP large model. By introducing the visual saliency calculation module and the visual block fine-grained selection integration module, the problems of multi-scale targets and redundant information in the remote sensing image are effectively solved, fine-grained semantic alignment between the image and the text is realized, and the retrieval accuracy is improved. Particularly, under the condition that the image contains a plurality of salient targets and redundant regions, the cross-modal semantic alignment fine-grained filtering method provided by the invention can accurately identify key information blocks in the image and perform fine matching with text description.

The invention provides an on-site video return coding method and system, and the method comprises the steps: collecting a video stream of an on-site scene, judging whether a current video frame in the video stream responds to a scene switching video frame or not, and marking the current video frame as a key frame when the current video frame responds to the scene switching video frame; after the current video frame is identified as a key frame, visual saliency analysis is performed on the key frame, and an entropy weight matrix representing regional information importance distribution in the key frame is constructed according to a visual saliency analysis result and texture information entropies of different blocks in the key frame; adjusting code rate allocation weights of different blocks in the key frame according to the entropy weight matrix and a code rate regulation and control strategy of the key frame to obtain adaptive coding configuration adaptive to content characteristics of the key frame; and performing optimization coding on the key frame through adaptive coding configuration to obtain a target code stream in response to a field video return demand. By adopting the scheme of the invention, the dynamic differential coding of the key video frame in a complex scene can be realized.

The invention relates to a power distribution network tower defect automatic identification and classification method and system based on deep learning. According to the method, firstly, a tower main body area is positioned and extracted through a convolutional neural network, and background interference is eliminated; and then the visual saliency of the defect area is improved by adopting a self-adaptive contrast enhancement algorithm based on local statistical characteristics. In the feature extraction stage, a pyramid distraction attention module is introduced to fuse multi-scale space information and channel attention, and a two-dimensional selective state space module is used for modeling a long-range dependency relationship. And multi-resolution features are further aggregated through a layered feature fusion architecture and a self-adaptive anchor frame mechanism, and targets of different sizes are matched. And finally, a self-adaptive edge enhancement module is adopted to enhance the edge of the defect, and a multi-branch detection head is adopted to realize category judgment, position regression and confidence evaluation of the defect in parallel. The method effectively improves the detection precision and robustness of tower defects under a complex background, and is especially suitable for the automatic recognition of micro-scale defects.

The invention discloses a hyperspectral anomaly detection method and system for enhancing low rank and significance. The method comprises the following steps: acquiring hyperspectral image tensor data of a to-be-detected region; carrying out band-by-band normalization processing; decomposing into a background tensor and an abnormal tensor; designing an improved weighted tensor nuclear norm based on tensor singular value decomposition; constructing a visual saliency sparse weight tensor; establishing an anomaly detection model; obtaining an optimal abnormal tensor; and detecting the obtained optimal abnormal tensor to generate an abnormal detection result graph. According to the method, comprehensive mining of background low-rank features and full utilization of abnormal target saliency can be realized, and the detection rate of hyperspectral anomaly detection is effectively improved; the abnormal target can be clearly and accurately detected, the shape of the abnormal target can be completely reserved, good detection performance is still kept under the complex background and noise interference, and reliable technical support is provided for related applications such as mineral exploration and military reconnaissance.

The invention relates to the technical field of display screen layout control, and discloses an AI-based display screen layout control method and system, and the method comprises the steps: obtaining the operation record and focus distribution data of a user, and obtaining the time sequence data of a user behavior; extracting dynamic change rule characteristics of user behaviors according to the time sequence data; user interaction records are mined based on dynamic change rule features, classification is carried out according to time and scene dimensions, and long-term trend features and short-term fluctuation features of user behaviors are determined; obtaining the attention weight value of each layout element, and determining a visual saliency sequence; priority values of high-priority elements are calculated in combination with visual saliency sorting, and a preliminary layout scheme is obtained; performing amplitude adjustment according to the preliminary layout scheme to obtain an adjusted layout scheme, and adaptively changing the sizes and positions of layout elements to obtain an optimized display screen layout. According to the method, the display screen interface layout can be dynamically optimized, and the user interaction experience and the working efficiency are improved.

A computer system displays a representation of a user pose in a three-dimensional environment in response to movement of a user's current viewpoint. The computer system displays different representations of the movement of the virtual representation based on the type of the virtual representation of the user. A computer system reduces visual saliency of a virtual representation when changing a spatial arrangement of virtual objects shared in a communication session. The computer system displays different visual feedback when moving the virtual object depending on whether the virtual object is shared or not shared in the communication session. The computer system displays visual feedback indicative of audio provided by another user. The computer system displays feedback indicating that the participant will correspond to the location. The computer system displays a sequence of visual transitions while displaying a visual representation of a participant in the communication session.

This invention relates to the fields of remote sensing image analysis and cross-modal retrieval technology. It discloses a remote sensing image text retrieval method based on a remote sensing multimodal fundamental model. By fine-tuning the CLIP large model, its large-scale pre-training capabilities are applied to the semantic alignment of remote sensing images and text. By introducing a visual saliency calculation module and a visual block fine-grained selection integration module, this invention effectively solves the problem of multi-scale targets and redundant information in remote sensing images, achieving fine-grained semantic alignment between images and text, and improving retrieval accuracy. Especially when images contain multiple salient targets and redundant regions, the proposed cross-modal semantic alignment fine-grained filtering method can accurately identify key information blocks in the image and perform fine matching with the text description.

The invention discloses a character animation key frame simplification method and system based on visual saliency, and relates to the technical field of image processing, and the method comprises the steps: obtaining a character animation sequence, and extracting a position path and a rotation path of each skeleton node in the character animation sequence in a time dimension; calculating a first motion feature sequence based on the position path of each skeleton node; calculating a second motion feature sequence based on the rotation path of each skeleton node; combining the first motion feature sequence and the second motion feature sequence of each skeleton node into a third motion feature sequence; and selecting a target key frame sequence from the original time points based on the third motion feature sequences of all skeleton nodes by taking minimization of a motion reconstruction error of the role animation sequence as an optimization target, and outputting the target key frame sequence. Through feature extraction and fusion combination of the position path and the rotation path, accurate screening of animation key frames is realized, redundant frames are reduced, and data compression efficiency and animation coherence are improved.

The application relates to the technical field of video recognition, and discloses a course video key frame intelligent recognition method based on an AI visual attention mechanism, which comprises the following steps: acquiring a visual saliency feature map of a video frame and calculating a global attention gravity center coordinate, constructing a spatial second moment tensor by using the visual saliency feature map to determine an anisotropy coefficient, then performing nonlinear weighted processing on the trajectory distribution density in a space-time trajectory space, calculating a second acceleration residual based on the processed trajectory evolution process, and determining a key frame in combination with the trajectory distribution density and the second acceleration residual. The application uses an anisotropy regulation mechanism to suppress non-content dynamic interference, checks the integrity of teaching content generation through the second acceleration residual, solves the problem of lag in semantic turning point capture under a dynamic background, and enhances the semantic density of extracted sequences and the recognition stability.

The application relates to the technical field of data processing, and discloses a subject Token generation system and method based on image feature extraction. The system integrates and divides semantic regions by hierarchically extracting and fusing multi-level visual features, obtaining a fused feature map, and providing basic feature data for subsequent semantic matching and feature quantization. According to domain subject knowledge, a subject semantic structure is built, and a quantization codebook is initialized. A mapping relationship between codebook entries and subject semantics is established, so that the codebook is self-adaptive to different image semantic scenes. In the attention interaction link, the mapping relationship is introduced to impose semantic constraints, and a sparse subject Token set is generated, thereby improving the authenticity and effectiveness of subject representation content. The saliency score of the subject Token is calculated by fusing the visual saliency and attention weight of the semantic region. After global deduplication and sorting, a subject Token sequence is formed, which comprehensively and objectively reflects the core subject information of the image.

The invention discloses an image storage method, image storage equipment and a computer storage medium, and relates to the technical field of image processing and storage. The method comprises the following steps: acquiring original data through an image acquisition sensor, calculating local contrast and edge density, performing fusion to generate a visual saliency map, and performing region segmentation to obtain an average visual saliency value of each region; calculating a statistical variance and a global peak value according to the values of all the areas, and generating an adaptive rate distortion reference value; comparing each region value with a reference value to determine a quantization step size and a visual saliency level, and performing adaptive coding on the region to form a composite data stream; calculating the proportion of the geometric center to the area of each region, packaging the proportion into metadata, and packaging and storing the metadata and the coding stream; and constructing a mapping index from the region identifier to the physical address and a reverse index based on the visual saliency level. The method can improve the utilization rate of the storage space and optimize the image quality of the important region.

The application discloses a kind of based on group intelligence optimization's ciphertext domain image visible watermark embedding method and system, belong to the field of encrypted storage technology.The method includes obtaining original plaintext image, using preset encryption algorithm to the original plaintext image is encrypted, generate ciphertext image, to ciphertext image is remaining entropy space modeling, and according to copyright information generates visible watermark image;Watermark embedding parameter set to be optimized is constructed, watermark embedding parameter set is encoded as individual position vector in group intelligence algorithm, executes the multi-objective optimization search based on group intelligence, carries out optimal parameter selection set ciphertext domain watermark embedding based on search result, obtains the ciphertext image with visible watermark, with visible watermark ciphertext image is safely distributed and service processing handle.The application directly embeds visible watermark in encrypted image, and automatically weighs security, visual saliency and function reservation by group intelligence optimization, realizes the unification of security and intuitive copyright declaration.

The invention relates to the technical field of model loading, in particular to a 3D model dynamic loading method using a Unity engine, and the method comprises the following steps: obtaining an OSGB file of a 3D model obtained by image data reconstruction, converting the format, importing the OSGB file into the Unity engine, and constructing a tree index structure by using a file naming rule, each node is associated with the corresponding grid model and the mapping image; when the Unity engine runs, all nodes located in the view field of the main camera at all moments are obtained; and calculating the visual field fading degree, the visual richness, the visual saliency, the motion coefficient and the rendering evaluation value of each node, optimizing the node evaluation function of the LOD algorithm, and dynamically scheduling and loading the grid model and the mapping image of the corresponding detail level according to the evaluation result of the node for rendering. According to the method, high-performance rendering is ensured, and meanwhile, the reality sense and immersion sense of a scene picture are improved.

The application discloses a role animation key frame simplification method and system based on visual saliency, and relates to the technical field of image processing, comprising: acquiring a role animation sequence, extracting the position path and rotation path of each bone node in the role animation sequence in the time dimension; calculating a first motion feature sequence based on the position path of each bone node; calculating a second motion feature sequence based on the rotation path of each bone node; combining the first motion feature sequence and the second motion feature sequence of each bone node into a third motion feature sequence; and selecting a target key frame sequence from the original time points based on the third motion feature sequence of all bone nodes, with the optimization target being to minimize the motion reconstruction error of the role animation sequence, and outputting the target key frame sequence. Through the feature extraction and fusion combination of the position path and the rotation path, the application realizes accurate screening of animation key frames, reduces redundant frames, and improves data compression efficiency and animation coherence.

The application provides a first-view gaze point prediction method based on attention shift, comprising: extracting an optical flow image; constructing a first-view gaze point prediction model based on attention shift; inputting the optical flow image into the first-view gaze point prediction model to obtain spatial features and time features; obtaining an initial visual saliency image and an attention image based on the spatial features and the time features; and fusing the initial visual saliency image and the attention image to generate a final gaze point prediction image. The application extracts spatial and temporal features of an original image at multiple scales, fully utilizes time sequence information and high-level semantic information for saliency detection, predicts subsequent attention from previous gaze by modeling attention shift, and finally combines a visual saliency model to fuse into a final gaze point prediction image, thereby improving the accuracy of gaze point prediction.

The invention relates to the technical field of artificial intelligence, and provides an image generation method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the saliency analysis of an original image, obtaining the visual saliency information of the original image, and enabling the visual saliency information to represent the saliency of a visual target in the original image; determining transformation parameters of the original image based on the visual saliency information of the original image; and performing geometric transformation on the original image based on the transformation parameters to obtain a generated image. According to the method, the device, the electronic equipment and the storage medium provided by the invention, the transformation parameters are determined based on the visual saliency information of the original image, and geometric transformation is performed on the original image, so that more flexible single-image multi-view enhancement can be realized while target feature damage is avoided, the calculation cost is reduced, and the user experience is improved. The method can effectively support scenes, such as data enhancement, image registration and the like, which need to carry out diversified processing on a single image, and improves the generalization ability of the model.

The invention provides an automatic generation method, system and device of an OTA shop business card and a storage medium, and the method comprises the steps: receiving an image related to a shop and uploaded by a merchant, analyzing the image related to the shop through an image processing algorithm, dividing the image into a plurality of regions through a visual saliency detection algorithm, and storing the regions in a database; performing color clustering in a predetermined color space by using a clustering algorithm in the at least one divided region so as to extract at least one dominant hue in the region; and generating at least one background color matching scheme by applying a preset color theory on the basis of the at least one dominant hue extracted in the intelligent color selection analysis step, applying the background color matching scheme selected by the user in preview display to a corresponding area of a preset shop business card template, and generating and outputting the shop business card applying the selected background color matching scheme. According to the method, the automation degree, visual consistency and generation efficiency of business card design can be improved.

This application discloses a method, apparatus, and electronic device for visual content identification, processing, and tracing. The method includes: acquiring first visual content uploaded by a user and encoding the user identifier to obtain encoded information; performing a frequency domain transformation on the first visual content to obtain a first frequency domain coefficient matrix; modifying at least some frequency domain coefficients corresponding to the target frequency domain region based on the visual saliency features and encoded information corresponding to the target frequency domain region in the first frequency domain coefficient matrix to obtain a second frequency domain coefficient matrix embedding encoded information; generating second visual content based on the second frequency domain coefficient matrix; and when the second visual content is determined to be in violation by the server, quickly reconstructing the user identifier from the infringing image, rapidly tracing the infringing user account, and taking timely action.

The present application relates to the technical field of image analysis, in particular to a beef cattle intake intelligent monitoring and analysis method based on video stream, comprising the following steps: converting single-frame beef cattle shed RGB images in the video stream to logarithmic chrominance space.In the present application, by performing chrominance space conversion on single-frame beef cattle shed RGB images and separating illumination and reflection components, shadow interference can be suppressed, the feeding trough area in the image is more stable, the shadow suppressed feeding trough image is divided into overlapping blocks, local structure features and visual saliency features are calculated to generate a priority matrix, so that the local enhancement process can be adjusted according to the actual feature distribution rather than a single global index, the flexibility and accuracy of the local area in contrast adjustment are improved, and combined with the enhanced priority matrix, adaptive contrast limit amplitude is set for different blocks, the distinction between the feed area and the background is improved, and the texture and boundary features in the area are more prominent.

The invention discloses a domain-division dynamic adaptation smooth transition watermark embedding method and system based on image saliency. The method comprises the following steps: acquiring an original visual image; acquiring a visual saliency map according to the original visual image; dividing the visual saliency map into a high saliency region, a transition region and a low saliency region; on the basis of different regions, a Gaussian smooth weight and an Euclidean distance field weight are fused, and a smooth transition weight map is constructed; and based on the smooth transition weight map, in combination with a dynamic regional embedding strategy, obtaining an embedding result. According to the method, the problems of rough region division, stiff boundary transition, insufficient local feature utilization, single low-saliency region processing and the like in the existing digital watermarking technology can be solved.

The invention discloses a myopia prevention method based on a neural network model, particularly relates to the technical field of visual health monitoring, and is used for solving the problems that the real-time dynamic visual physiological load of a user cannot be perceived and responded due to dependence on a static rule and personalized myopia risk early warning and prevention are difficult to realize in the prior art. Real-time visual behavior data of a user in immersive experience equipment is acquired, a visual adjustment compensation state of the user is evaluated based on a neural network model, and when the state exceeds a preset threshold value, a visual behavior entropy is further calculated and a content semantic focus change rate of a virtual scene is synchronously analyzed so as to analyze a dominant inducement type of state abnormality; matching the fixation point distribution with the scene visual saliency topological graph according to the inducement type to identify specific dominant factors, and dynamically adjusting display parameters of the immersive experience equipment according to the specific dominant factors; accurate evaluation and inducement analysis of the real-time load of the user visual system are achieved, and the myopia prevention effect is achieved while the immersion experience is maintained.

The application discloses a kind of self-media promotion content heat prediction methods based on deep learning, comprising: S1, acquisition multi-modal data and pre-process output text word vector sequence and standard size image;S2, construct semantic structure atlas and output text semantic feature vector;S3, carry out saliency target detection and output visual saliency feature map;S4, feature aggregation and interaction are carried out using MLP-Mixer model, and output graphic-text coupling feature vector;S5, construct improved DCCA network, introduce covariance reconstruction loss function to generate psychological resonance index feature;S6, carry out time series evolution prediction and output heat prediction result.The application realizes graphic-text feature deep coupling and audience psychological resonance quantization, effectively improves the accuracy and explainability of heat prediction.

The application discloses a kind of based on multimodal fusion's eye movement tracking method and system, specifically related to eye movement tracking technical field, including through the RGB camera, infrared camera and IMU of synchronous setting, user face image, eye image and head movement parameter are collected;Parallel processing these signals, respectively extract the line-of-sight direction vector based on face and eye and its reliability measure, and solve head space posture;According to reliability measure dynamic selection high confidence cooperation, master-slave compensation or conflict arbitration strategy, fusion generates initial line-of-sight direction;The direction is mapped to display plane to obtain initial gaze coordinate, and optimization calibration is carried out in combination with the visual saliency analysis of screen content, and the final gaze point coordinate is output.The system includes signal acquisition processing, parallel computing, multi-strategy fusion decision and output optimization module.The application significantly improves the accuracy and robustness of eye movement tracking through multimodal information fusion and dynamic strategy selection.

The invention relates to the technical field of video recognition, and discloses a course video key frame intelligent recognition method based on an AI visual attention mechanism, which comprises the following steps: acquiring a visual saliency feature map of a video frame, calculating a global attention barycentric coordinate, constructing a spatial second moment tensor by using the visual saliency feature map to determine an anisotropy coefficient, and calculating an anisotropy coefficient of the video frame; the method comprises the following steps of: performing nonlinear weighting processing on a trajectory distribution density in a space-time trajectory space, calculating a second-order acceleration residual error based on a processed trajectory evolution process, and determining a key frame by combining the trajectory distribution density and the second-order acceleration residual error. The integrity of teaching content generation is verified through the second-order acceleration residual error, the problem of semantic turning capture lagging under the dynamic background is solved, and the semantic density of the extracted sequence and the recognition stability of the extracted sequence are enhanced.

A method of processing an image, an electronic device, and a storage medium. The method includes: determining, based on a predetermined color mode, a second rendering color of pixels of an original image according to a first rendering color of the pixels, where the first rendering color corresponds to a first color space; adjusting the second rendering color according to the second rendering color and an original rendering color of the pixels to obtain a third rendering color of the pixels, the original, second, and third rendering colors corresponding to the second color space, the first rendering color being obtained by performing a color space conversion on the original rendering color; and obtaining a target image according to the third rendering color, where a region corresponding to a visual saliency region of the original image in the target image has a color corresponding to the predetermined color mode.

The application discloses a kind of intelligent behavior detection method and system based on motion feature enhancement, construct based on adaptive mixture Gaussian model, the Mahalanobis distance of pixel point and background distribution is calculated and model weight is dynamically updated, realize the accurate decoupling and separation of foreground motion region in video frame;Then, the foreground mask obtained initially is based on the connected domain analysis of area threshold to filter out discrete noise, and the morphological closing operation of structure element adaptation is used to fill target internal cavity and repair edge fracture, to generate the high signal-to-noise ratio of motion region binary graph;Again, the purified binary graph is mapped as high visual saliency color heat map, it is weighted linearly fused with original RGB video frame, the fusion feature map is used to guide deep learning model to pay attention to motion region, and through double threshold detection strategy, the perception and identification ability of system to small target and hidden uncivilized behavior under complex dynamic scene is greatly improved.

The application relates to the technical field of display module dynamic refresh rate adjustment, and particularly discloses a display module dynamic refresh rate adjustment method based on content sensing, which extracts pixel update features and visual saliency weights of a to-be-rendered layer in a graphics rendering pipeline, generates content dynamic characteristic values to represent content dynamic attributes, dynamically modulates a physical clock frequency of a transmission interface according to the content dynamic characteristic values, establishes a variable-frequency data transmission link, matches a data transmission bandwidth with content load, generates elastic scanning control instructions according to a real-time data transmission rate, adjusts a row scanning period and a gate driving voltage, forms a dynamic load row scanning timing, and when a high-priority touch event is detected, forcibly resets an association state of data transmission and scanning control, and ensures instant responsiveness of user interaction. Through fine content analysis and dynamic refresh rate adjustment, the application effectively reduces display module power consumption.

The invention relates to the technical field of defect detection, in particular to an off-highway wide-body mining vehicle axle surface defect detection system which comprises a multi-source data geometric registration module, a fusion mapping module, a visual saliency enhancement module, a defect area positioning module, a defect classification and quantification module and a detection report generation module. Pixel-level geometric registration is carried out on the texture image data and the light point cloud data of the axle surface; determining a color three-dimensional point cloud image; carrying out gradient fusion enhancement on the color three-dimensional point cloud image to obtain a visual saliency image; performing superpixel clustering on the surface of the axle, and performing spatial outlier identification on a result after clustering to obtain a defect area; judging the defect type of the axle surface, and determining defect quantitative description information in combination with the texture spectrum; integrating the defect area, the defect category and the defect quantitative description information to obtain a defect detection report; according to the invention, the efficiency of detecting the surface defects of the axle of the off-highway wide-body mining vehicle can be improved.

The invention belongs to the technical field of vegetation ecological monitoring, and particularly relates to a multi-source fusion vegetation ecological monitoring method and system, and the method comprises the steps: synchronously collecting environment data and visual data in a monitoring region; calculating the fluctuation amplitude of the environment data in the time sliding window, and constructing an environment fluctuation potential energy index for representing the instability of the external physical environment based on the fluctuation amplitude; calculating the difference between the current frame image and the reference frame image according to the visual data, and constructing a visual saliency difference entropy in combination with the image brightness information; and carrying out weighted calculation on the visual saliency difference entropy by utilizing an environmental fluctuation potential energy index, generating a self-adaptive return priority coefficient, and executing a corresponding data transmission strategy. According to the invention, invalid visual change is inhibited through physical environment fluctuation, the number of times of invalid communication is effectively reduced, and low-power-consumption monitoring in a normal state and reliable early warning under extreme disasters are realized.

The application discloses a ground subsidence funnel identification method, system, device and storage medium, and relates to the field of surveying and mapping technology. The method comprises the following steps: loading a single-band deformation gray result map of synthetic aperture radar interferometry and converting the single-band deformation gray result map into enhanced RGB visual features; constructing a diversified interpretation sample set for model training; performing supervised training on the improved network model by using the constructed sample set; deploying the trained model in actual business, inputting an InSAR deformation monitoring single-band gray map to be identified into the trained model, and outputting a ground subsidence funnel identification result. The single-band gray deformation result object is converted into a color-enhanced image, and the visual saliency of the subsidence funnel is strengthened. The sample augmentation method is applied, the robust generalization ability of the deep learning model is enhanced, the improved model is introduced, the precise perception and segmentation ability of the network for subsidence funnels of different scales and shapes are improved, and the background and artifact interference is effectively inhibited.