32 results about "Image object" patented technology

This application relates to the field of medical technology and provides a medical application scenario matching method, electronic device, and computer program product. The medical application scenario matching method includes: acquiring a medical image sequence; determining image information of the medical images in the medical image sequence, the image information including key information, which includes one or more of the following: imaging object information, phase information, lesion detection information, and image quality information of the corresponding medical image; and outputting at least one target application scenario adapted to the medical image sequence based on the key information. Embodiments of this application can prevent doctors from using unsuitable medical image sequences in specific application scenarios, thus helping to improve doctors' work efficiency.

This invention discloses a user-generated content sharing system with fine-grained access control, specifically relating to the field of user-generated content sharing and data access control. It includes an end-to-end sharing module for writing published user-generated content into fragment identifiers based on text segments, image objects, and attachment objects, attaching the initial receiving scope and initial re-sharing action to the fragment identifiers, and generating an original shared copy. This invention solves the problem in existing technologies where, when receiving user-generated content, it is impossible to limit the content segments that can be brought out, the receiving object scope, and the re-sharing action based on the current shared copy by splitting user-generated content into fragments and generating an original shared copy.

An information processing device (100) comprises: an acquiring unit (121) that acquires image information representing a two-dimensional image; a determining unit (122) that determines a reference object to serve as a reference from among one or more image objects included in the two-dimensional image represented by the acquired image information, and, on the basis of the dimensions of the determined reference object, determines display information representing a display region in which the two-dimensional image is displayed in a virtual space; and an output unit (123) that outputs the determined display information.

An image object labeling method and circuit system for implementing the method, in which an image is first obtained, the image is divided into one or more blocks by object classification, each block is classified into a category and is assigned a category label, and depth information of each pixel in the image is estimated by a depth estimation method. Then, according to the category of each block, it is determined whether the depth information of each pixel in the block matches the category of the block to which the pixel belongs. When the depth information of any pixel in each block matches the category of the block to which the pixel belongs, weights are set according to the category labels of each block, and a post-processing procedure is performed on the image according to the weights of each block. Otherwise, if the depth information of any pixel does not match the category of the block to which the pixel belongs, it is considered as noise and the post-processing procedure is not used.

This invention discloses a remote sensing classification method and device for periglacial landforms. The method includes: establishing remote sensing classification rules for periglacial landforms, dividing periglacial landforms into primary and secondary periglacial landform types; collecting remote sensing images and digital elevation model data of the target area in permafrost regions at various resolutions, and performing radiometric calibration and atmospheric correction on the remote sensing images; performing first-scale segmentation and second-scale segmentation on the processed data to generate two-level image object layers, with the first-scale segmentation generating a primary image object layer corresponding to the primary periglacial landform type and the second-scale segmentation generating a secondary image object layer corresponding to the secondary periglacial landform type; extracting and analyzing features for each image object in the primary and secondary image object layers; and, based on the remote sensing classification rules for periglacial landforms, classifying the image objects in the primary and secondary image object layers into the corresponding primary and secondary periglacial landform types, respectively, and outputting a remote sensing classification map of periglacial landforms.

Provided are a method for generating a bounding box of an image target, an electronic device, a non-transitory computer-readable storage medium, and a computer program product. The method includes: obtaining a feature map based on one or more of a self-attention feature, a cross-attention feature, and a latent vector obtained in a process of generating an image including an image target or in a process of adding noise to the image and regenerating the image; and obtaining a bounding box of the image target based on the feature map.

A method for manufacturing a synthetic image device includes providing of a focusing element array. An image layer is arranged in a vicinity of a focal distance of focusing elements, whereby a synthetic image composed of enlarged portions of the image layer becomes perceivable for a viewer. The image layer includes an array of image cells, each associated with a respective focusing element. The step of arranging the image layer includes creation of a respective image object within each respective one of said image cells. The image objects are such that an animation becomes perceivable, comprising a series of synthetic images perceivable in-series as the viewing direction changes. The image objects are such that each one of the synthetic images of the series is perceivable at a respective perceivable depth, changing between the synthetic images of the series of synthetic images.

The application relates to the field of medical image archiving, and discloses a sequential medium medical image archiving method, system, terminal and storage medium based on visibility root and reachability recovery control, which comprises the following steps: aggregating medical image objects into immutable archive segments and sequentially writing the medical image objects into a target area; after the writing is completed, a release record containing a segment identifier, an area identifier and a predecessor identifier is added to a release log; an atomic update of a visibility root pointing to the release record is independently controlled on a storage layer, the visibility root is a unique entry of a logical readable state; and all reachable release records are determined along an ancestor chain formed by the predecessor identifier, starting from the visibility root, and the corresponding immutable archive segments are external readable archive segments, so that archiving is completed. The physical writing completion is decoupled from the logical visibility release, the read visibility range is restricted through the visibility root, and the write area recovery condition is controlled, so that consistent readability, recoverability and controllable recovery of medical image archiving on a sequential medium are realized.

Devices, systems, and techniques for using one or more neural networks to map object features are described. In at least one embodiment, one or more neural networks are used to generate one or more image object features, based at least partially on a variable sampling rate of one or more audio signals.

An image search method includes: providing a human-machine interface to receive an input of a user; performing an image object detection operation according to the input of the user sent from the human-machine interface to generate a specific image; performing an image-to-semantic operation based on the specific image to perform an irreversible encryption encoding upon the specific image to generate a specific semantic vector; sending the generated specific semantic vector into a cloud semantic search engine of a cloud server so that the cloud semantic search engine searches multiple semantic vectors stored in the cloud server according to the specific semantic vector so as to select image information corresponding to an approximate semantic vector and send the image information into the human-machine interface; and, controlling the human-machine interface to display the image information for the user.

Embodiments of the present application disclose a radar image object detection method, device and equipment and a storage medium. The method comprises: acquiring a radar sample image, and determining initial detection contribution values of different depth information in the radar sample image to a to-be-detected object; inputting the radar sample image into a to-be-trained deep learning model, introducing an attention mechanism in the to-be-trained deep learning model according to the initial detection contribution values, and performing object detection on the radar sample image; adjusting model parameters of the to-be-trained deep learning model and the initial detection contribution values according to a loss function of object detection of the to-be-trained deep learning model, to obtain a target deep learning model and target detection contribution values; and performing object detection on a target radar image by using the target deep learning model and the target detection contribution values. The method introduces an attention mechanism for different depth information in a radar image by using detection contribution values, and can improve the object detection accuracy of the radar image.

The application relates to the technical field of image target detection, and particularly relates to a weakly supervised remote sensing image target detection method based on multi-modal pseudo label guidance and adaptive fusion. Through a cross-modal adaptive relationship fusion module, double-modal features are dynamically fused; and through a cross-modal pseudo supervision label mapping module, RGB and depth pseudo labels generated and aggregated are used, a unified structure loss and a depth double loss are proposed to supervise and train a main branch and an auxiliary branch, so as to solve problems such as insufficient utilization of multi-modal information, large pseudo label noise and insufficient global and local information fusion under weak supervision.

The application discloses a data processing method and device, electronic equipment and storage medium, comprising: reading file data of a read-only file, the read-only file comprising an image; based on the file structure of the file data, finding the storage location of the image object; according to the storage location, extracting the image object from the file data; data decoding the image object to obtain image data; target detection is performed on the image data to obtain a target image comprising a target object; for each target image, based on the received editing processing requirement, the target image is edited and processed to obtain a new image; the new image is used to replace the target image corresponding to the storage location in the read-only file to obtain a new read-only file. The data processing method provided in the embodiment of the application can realize the editing processing process of the image in the read-only file only by inputting the editing processing requirement of the user, improve the efficiency of data processing, and greatly reduce the image processing threshold.

Disclosed embodiments may provide techniques for generating a differentiable mesh representation of an image object. A computer-implemented method can include accessing a set of points, in which each point is assigned with a weighted value. The method can also include receiving ground-truth input to generate a differentiable mesh representation associated a target object. The method can also include constructing an initial mesh representation of the target object by generating a plurality of simplices based on the weighted values of the points. The method can also include optimizing the initial mesh representation by iteratively adjusting the existence probability values based on a comparison between the initial mesh representation with the ground-truth input. The method can also include extracting the differentiable mesh representation of the target object by selecting a set of simplices of the plurality of simplices associated with the initial mesh representation.

This invention discloses a method for extracting battery material data based on multimodal collaboration and context completion, belonging to the field of battery material data extraction technology. The method includes: acquiring text object datasets and image object datasets; acquiring a global context feature dictionary; acquiring an image extraction dataset; performing dynamic logical completion and confidence scoring; performing multimodal data fusion and conflict resolution to obtain the multimodal data fusion result; performing data verification and standardized output to obtain a battery material performance knowledge base. This invention solves the technical problems of existing technologies, such as isolated multimodal data extraction from battery material literature, inability to complete missing information, and lack of standardized extraction processes, leading to low extraction efficiency, poor data integrity, and insufficient accuracy. It achieves collaborative extraction, intelligent completion, and standardized output of multimodal data from battery material literature, improving the efficiency, completeness, and accuracy of battery material data extraction.

The application discloses a semi-supervised medical image target detection method based on DETR, first, through the deblurring processing of the image, the inaccurate blur kernel and the additive Gaussian noise are used to model the image degradation, and a variety of prior knowledge and optimization algorithms are combined to restore the clear image. Secondly, a pair of teacher and student models with the same network architecture are adopted, and the strong and weak enhanced versions of the unlabeled image are used for training, and the training of the student model is guided by the pseudo label of the teacher model. At the same time, the cost-based pseudo label mining module dynamically mines the reliable pseudo bounding box in the unlabeled data, which is used for cross-view query consistency calculation. Finally, the supervised loss, unsupervised loss and cross-view consistency loss are combined with the corresponding weights to guide the training of the model. The method is simple and efficient, and can adapt to complex and changeable medical image structures, and provides accurate target detection information for clinical diagnosis.

This application provides an electronic device and an image search method. One image search method includes: providing a human-computer interaction interface to receive user input; performing image object detection based on the user input transmitted from the human-computer interaction interface to generate a specific image; performing image-to-semantic conversion based on the specific image to irreversibly encrypt and encode the specific image to generate a specific semantic vector; sending the specific semantic vector to a cloud-based semantic search engine on a cloud server, causing the cloud-based semantic search engine to search multiple semantic vectors stored on the cloud server based on the specific semantic vector to select image information corresponding to approximate semantic vectors and send the image information corresponding to the approximate semantic vectors to the human-computer interaction interface; and controlling the human-computer interaction interface to display the image information to the user.

The aim is to provide effective support for object positioning. [Solution] An information processing method is provided, comprising imaging object processing, virtual object processing, and comparison processing, wherein the imaging object processing acquires an imaging object, the virtual object processing acquires a virtual object, and the comparison processing acquires a comparison result between the imaging object and the virtual object.

The application relates to an image target detection and instance segmentation method, system, computing device and medium, which comprises the following steps: inputting an image into a backbone residual network containing switchable hole convolution to extract features at different stages and obtain multi-scale feature maps at different depths; inputting the multi-scale feature maps into a neck network to perform recursive fusion of a recursive feature pyramid, obtain new feature maps, generate a candidate region and obtain a region of interest; and inputting the region of interest into a head network to predict an instance category, a category confidence, an instance bounding box and an instance mask. The application can effectively improve performance and effectiveness and has potential application value in a wider field. The application can be applied in the field of image data processing.

Apparatuses, systems, and techniques to use one or more nueral networks to image object features are described. In at least one embodiment, one or more neural networks are used to generate one or more image object features based, at least in part, on a varaible sample rate of one or more audio signals.

The present disclosure relates to an image processing method, an image processing apparatus, and a computer program product. The image processing method comprises: extracting a target image object from a first image position; in response to the target image object being moved to a second image position in the first image, adjusting a size of the target image object according to the second image position, the second image position being different from the first image position; and fusing the target image object with the first image.

A toy provided by the present utility model includes a connecting component (442) that connects two components (441, 443) forming the appearance of an image object such as a small robot toy (20, 30) in a manner that allows them to be separated from and contacted with each other. One end of the connecting component (442) is combined with one of the two components (441), and the other end is combined with the other of the two components (443) in a manner that can rotate on the same plane, and it has a connection structure that can maintain the displacement state of the other component (443) relative to the one component (441). With the toy of the present utility model, the components connected by one connecting component can move relatively freely in appearance and can maintain the state after movement.

A method for processing breast tissue image data includes processing the image data to generate a set of image slices collectively depicting the patient's breast; for each image slice, applying one or more filters associated with a plurality of multi-level feature modules, each configured to represent and recognize an assigned characteristic or feature of a high-dimensional object; generating at each multi-level feature module a feature map depicting regions of the image slice having the assigned feature; combining the feature maps generated from the plurality of multi-level feature modules into a combined image object map indicating a probability that the high-dimensional object is present at a particular location of the image slice; and creating a 2D synthesized image identifying one or more high-dimensional objects based at least in part on object maps generated for a plurality of image slices.

Method and arrangements for forming image data for use in 3D imaging of an object (520). The image data being from an imaging system (505) based on light triangulation and that provides one or more illuminations (511) in one or more light planes that geometrically correspond to the same plane or parallel planes. The imaging system (505) and camera(s) (530) thereof are configured to image object points of the object (520) at least twice, according to a first imaging and second imaging, during illumination of the object points by said illumination(s). The first imaging and second imaging differing in how specular reflections caused by said illumination(s) are being imaged. First and second image data from the first and second imaging are obtained, corresponding first and second profile images (670) with common mapping to real world coordinates for intensity peaks (682, 683) caused by direct reflections from the object (520). Image data for use in the 3D imaging is formed (802) based on the obtained image data and excludes intensity peak data that for a respective object point imaged twice have intensity peak characteristics that differ between the first and second image data.

A computer implemented method, system, and non-transitory computer-readable device that may be used in a remote deposit environment. A method may include activating a camera on a client device, wherein the activated camera provides access to a field of view of the camera;detecting, based on a Light Detection and Ranging (LIDAR) measurement, a first distance of a document surface, wherein the document surface is located at least partially within the field of view of the camera; selecting, based on the first distance, a focal length for the camera;setting the camera to the focal length; generating, by the camera set to the focal length, an image object of at least a portion of a first side of the document; and performing an optical character recognition (OCR) process on the image object to extract one or more data fields.

The application discloses a kind of dynamic self-adapting graph-based image target tracking and motion prediction method, comprising: image frame is collected and is cached;Target detection is carried out and semantic attribute is extracted;Constitute confidence-enhanced graph node;Sparse graph topology based on Delaunay triangulation algorithm is constructed;Confidence weighted node affinity is calculated;Edge affinity is calculated;Secondary integer programming matching model is constructed;Continuous relaxation and Sinkhorn projection solution are carried out;Hungarian discretization and optimal matching output are carried out;Adaptive backtracking and trajectory chain generation are carried out;Adaptive prediction order is determined and motion derivative is calculated;Taylor series extrapolation prediction is carried out and prediction node set is output.The method realizes high robustness improvement in trajectory association aspect.Introduce sparse graph structure based on Delaunay triangulation and confidence perception mechanism, and combine reverse matching strategy, effectively utilize the spatial relationship constraint of target group, significantly enhance the tracking stability under complex scenes such as occlusion, detection loss.

The application provides a kind of image object positioning method and system based on generation model, comprising: for the generation process of picture synthesized from noise and text prompt of generation model, utilize image feature extraction network to obtain the image feature in process;For a certain object class in text prompt, utilize text encoder to extract text feature;Using cross-modal feature alignment network, the text feature and image feature are aligned, and the aligned representation is obtained;Using multilayer perception network, the aligned representation is mapped into image mask.The application can utilize the feature representation in natural image generation process, obtain the positioning information of object by aligning with the text feature of corresponding object class, and obtain the mask of corresponding object while generating picture.

An electronic device according to an embodiment may include a display, a memory for storing instructions, and a processor. When executed by the processor, the instructions may cause the electronic device to: display a still image including one or more image objects corresponding to text via the display; obtain text information corresponding to one or more image objects in the still image based on input for analyzing the still image being displayed; select a specified application corresponding to the still image from a plurality of applications using at least a portion of the text information; display a screen via the display including the still image and a visual object for launching the specified application; and display a launch screen for the specified application via the display in response to input associated with the visual object.

This application belongs to the field of image processing technology, specifically a method and system for intelligent synchronization of electronic photo frame images based on a local area network. The method includes: acquiring light and sound data, extracting light fluctuation frequency and average sound amplitude, and determining atmosphere type labels; then screening candidate images from local and online image resources, determining an atmosphere matching score by combining theme tags, color saturation values, and image object density values, and constructing a target image subset; subsequently generating and synchronously displaying a sequence, adjusting the carousel interval according to sound changes, and reconstructing the display sequence when changes exceed a threshold. This application can improve the environmental adaptability, content coordination, and multi-terminal consistency of image display.

Camera (140) registers image data (D) representing body parts (131, 132, 133, 134) of the milking animal (130). img At least one first frame. Image data (D) is registered using default exposure and / or ISO settings (xs). img Based on this image data, the controller (100) controls the automatic milking device (110) relative to body parts (131, 132, 133, 134). The controller (100) searches the image data (D) via a search process. img At least one region of interest is defined in at least one first frame of the image data. The search process is configured to detect at least one image object that satisfies at least one size-shape criterion, and the search process is performed in 3D image data. At least one region of interest is defined within the at least one image object that satisfies the at least one size-shape criterion. The controller (100) checks the data in each region of interest within the at least one region of interest to see if it satisfies a quality criterion. If the quality criterion is satisfied, the camera (140) continues to register the image data (D) using the default exposure and / or ISO settings (xs). img If the quality criteria are not met, the controller (100) controls the camera (140) to adjust its exposure and / or ISO settings (xs') so that the subsequently registered image data (D) img It is estimated to meet the quality criteria.