24 results about "Visible object" patented technology

The invention discloses a rendering method, which comprises the following steps: constructing an octree, and storing a to-be-rendered object in a to-be-rendered three-dimensional scene in a child node of the octree to obtain a space management octree; dividing the view cone into different view cone levels; traversing the space management octree, and determining a first intersection part between the space of the child node and the view cone and a view cone level to which the first intersection part belongs; distributing the first intersection part to different rendering queues according to the view cone hierarchy to which the first intersection part belongs; shielding and removing the first intersection parts in the different rendering queues to obtain current visible objects to be rendered; and rendering the current visible to-be-rendered object. According to the method, the octree is constructed to manage the three-dimensional scene space, the view cone levels are divided, the rendering queues are distributed, the view cone removing and multi-level shielding removing technologies are combined, the visible objects are accurately judged, the rendering load is effectively reduced, and the method has the advantages that the rendering performance is improved, space management is optimized, and the complex scene processing efficiency is improved.

The invention relates to the technical field of image rendering, and discloses a three-dimensional scene rendering method, device and equipment based on a graphics processor and a medium, the method is applied to the graphics processor, and the method comprises the following steps: receiving instance attribute data sent by a central processing unit; calling a calculation shader to distribute a calculation thread group according to the number of objects in the instance attribute data; calculating a corresponding transformation matrix for each object in parallel through the calculation thread group; adding the instance attribute data and the transformation matrix into an instance buffer area to obtain to-be-rendered instance data of each object; extracting to-be-rendered instance data from the instance buffer area, and performing invisible object removal processing on the to-be-rendered instance data through the calculation thread group to obtain target instance data of a visible object; and rendering the target instance data. According to the invention, the rendering effect and rendering efficiency of the three-dimensional scene are improved.

The invention discloses a scene data scheduling and rendering method based on dual PVS, and relates to the technical field of data processing. According to the method, the target scene unit where the camera viewpoint is located is determined, the rendering scene tree corresponding to the target scene unit is obtained, the nodes of the rendering scene tree are the target visible objects corresponding to the target scene unit, and then the view cone corresponding to the camera viewpoint is determined; and selecting a target node in the rendering scene tree as a target rendering object according to the view cone, determining a hierarchical detail model corresponding to the target rendering object according to the distance between the target rendering object and the camera viewpoint, loading the hierarchical detail model to the rendering pipeline from the rendering scene tree, and rendering the target rendering object. According to the method, the target visible objects are screened through the rendering scene tree, and then the target rendering object is determined in the target visible objects through the view cone, so that the objects needing to be rendered are further reduced, and the computing resource consumption of model rendering is reduced.

The present application provides for a video system that displays handheld objects in video capture by detecting objects being held by the user within a segmentation mask and adding the object to an allowed list of permissibly visible objects. The system may be configured, during live video capture, to show or hide selected objects. As an added advantage, such approaches are not dependent on whether the hand is holding the object and moving synchronously. For example, if a video application on a video system included a pen in the allowed list, and the user waves and rotates the pen such that the pen moves out of sync with the body of the user, the video system will still make the pen visible. Moreover, the video application provides an efficient object detection mechanism that detects the handheld object once the object is placed within the segmentation mask.

The invention discloses an image acquisition system of a camera and a method for constructing characteristics and positions of a visual object. The image acquisition system comprises an image acquisition system. The method has the beneficial effects that by utilizing the pixel change rule caused by the displacement of the camera and combining with the plumb line direction, the image region is dynamically divided, so that efficient and low-cost three-dimensional modeling is realized; a deep learning model depending on object attribute recognition is abandoned, and high convex, low concave and planar objects are quickly distinguished through direct comparison of pixel variation; a dynamic region division rule based on a camera attitude (such as an inclination angle) is designed to ensure that an image region can be effectively segmented in different motion states, and the dynamic adaptability is enhanced; through segmented focusing and lightweight pixel difference calculation, the data processing delay is reduced, and the real-time navigation requirement is met.

The application provides a 4D person-object interaction generation method based on a text description, which comprises the following steps: in stage one, 3D person-object interaction key frame recovery: a human body motion sequence is obtained through a human body motion model and uniformly down-sampled to extract a key frame; a human body grid is reconstructed through an SMPL-X model for each key frame to extract a vertex position and form a human body point cloud; an object position anchoring network takes the human body point cloud, an object template point cloud and a text prompt as input to predict an object position and generate a 3D person-object interaction key frame; in stage two, 4D person-object interaction sequence generation: a contact perception diffusion model is constructed to take a sparse 3D person-object interaction key frame as input, extract a conditional signal through a contact perception encoder; based on the conditional signal, the 3D person-object interaction key frame is subjected to time series interpolation through the contact perception diffusion model to generate a time series coherent dense 4D person-object interaction sequence. The application realizes natural and realistic 4D person-object interaction synthesis of an unseen object.

A raster-based method for generating non-planar reflections in real-time. The method is based on projecting each time all visible object's polygons on a single polygon of a non-planar reflecting surface. All the projections are transformed to camera view and moved through the raster pipeline creating a complete, mirrored image.

The present application provides a user interface control method and apparatus, a medium, and an electronic device. The present application is applicable to a monitoring program installed in a vehicle-end system. The method comprises: after a voice instruction is acquired, acquiring, from the voice instruction, operation description information of a target interface object in a user interface of any application process on a touch screen; by means of the operation description information, obtaining, from an operation information set, display information and operation action information of the target interface object mapped to the touch screen; and then, on the basis of the display information and the operation action information, performing an operation on the touch screen, and triggering the target interface object to perform a response corresponding to the operation. The monitoring program runs in the vehicle-end system in the form of a process, and is not coupled to other application processes of the vehicle-end system, thereby completely eliminating the need to intervene with the other application programs during development, enabling completely independent monitoring of the other application processes, improving the survivability of the monitoring program, and implementing the function of allowing every visible object to be converted into a spoken instruction.

This invention provides a method for determining what a driver should be looking at while driving. [Solution] In the method for determining the object to be seen, the driver's line of sight is detected (S11), the gaze area is detected based on the detected line of sight of the driver (S12), an image obtained by photographing the area around the vehicle is acquired (S13), a candidate object to be seen is detected by recognizing the object contained in the image (S14), a traffic scene is determined (S15), it is determined whether or not the candidate object to be seen corresponding to the determined traffic scene is included in the gaze area (S16), if the candidate object to be seen corresponding to the determined traffic scene is included in the gaze area, it is determined that the driver performed appropriate visual actions in the determined traffic scene (S17), and if the candidate object to be seen corresponding to the determined traffic scene is not included in the gaze area, it is determined that the driver did not perform appropriate visual actions in the determined traffic scene (S18).

The present application discloses a scene data scheduling and rendering method based on dual PVS, which relates to the field of data processing technology. The above method determines the target scene unit where the camera viewpoint is located, obtains the rendering scene tree corresponding to the target scene unit, and the nodes of the rendering scene tree are the target visible objects corresponding to the target scene unit. Then, the viewing cone corresponding to the camera viewpoint is determined, and the target node in the rendering scene tree is selected as the target rendering object according to the viewing cone. According to the distance between the target rendering object and the camera viewpoint, the level of detail model corresponding to the target rendering object is determined, and the level of detail model is loaded from the rendering scene tree to the rendering pipeline to render the target rendering object. The method first filters the target visible objects through the rendering scene tree, and then determines the target rendering object among the target visible objects through the viewing cone, further reducing the objects that need to be rendered and reducing the computing resource consumption of model rendering.

The present disclosure provides an object counting method, which relates to the field of artificial intelligence technology, specifically computer vision, image processing, deep learning and other technical fields, and can be applied to scenarios such as smart cities. The specific implementation scheme is as follows: by inputting the target object image to be processed into the object target detection model to perform object target detection, a plurality of prediction frames of visible objects, item classification information of the plurality of visible objects, and occlusion classification information of the plurality of visible objects are obtained; based on the prediction frames of the plurality of visible objects and the item classification information of the plurality of visible objects, the count of occluded objects under each item category is obtained; based on the prediction frames of the plurality of visible objects, the occlusion classification information of the plurality of visible objects, and the item classification information of the plurality of visible objects, the count of occluded objects under each item category is determined; and based on the visible object count under each item category and the count of occluded objects under each item category, the actual item count under each item category is obtained.

The present invention provides a method, apparatus, and electronic device for controlling the playback of sound effects. In response to a sound effect trigger event, the target object that triggers the sound effect and the target sound effect to be played are determined; and the visibility state of the target object is obtained; the visibility state indicates whether the target object can be seen by a controlled virtual object or other virtual objects; based on the visibility state of the target object and the position of the virtual camera corresponding to the controlled virtual object in the game scene, the playback mode of the target sound effect is determined, and the target sound effect is played; in response to a change in the visibility state of the target object, the playback mode of the target sound effect is updated, and the target sound effect is continued to be played based on the updated playback mode. In the above manner, the playback mode of the sound effect is determined based on the visibility state of the target object, and when the visibility changes, the playback mode of the sound effect is adjusted, so that the current player can hear the sound effects of the visible objects, thereby learning about the game situation of the entire game scene and improving the user experience.

Two-dimensional (2D) tracking data from multiple camera views is processed to determine three-dimensional (3D) object parameters. Image features are extracted from object data associated with 2D tracking across the camera views. The extracted features for objects visible in multiple views are aggregated and spatial representation tokens are generated based on the object data and reference vectors. Instead of computing birds-eye-view (BEV) features, 3D spatial parameters can be determined directly from the aggregated features and spatial representation tokens. Attention operations analyze relationships between different feature sets while frame-to-frame consistency is maintained in the determined 3D parameters. Multi-view feature aggregation and spatial token generation operate directly on 2D tracking data from surround-view cameras, allowing feature processing without view-specific geometric transformations. Direct 2D-to-3D transformation supports object tracking across different camera configurations and viewing conditions.

This invention discloses a hologram calculation and occlusion processing method based on shared object point sampling. Multiple pixels are selected as sampling center points on the hologram plane to form a sampling grid. For each sampling center point, a cone-shaped beam composed of several rays is emitted to perform object point sampling, filtering out visible object points. For each pixel on the holographic plane, K sampling center points are selected based on its positional relationship with the sampling center points. The union of the visible object point sets is calculated to determine the object point set corresponding to the pixel. Each object point in this set is traversed, and its complex amplitude contribution to the target pixel is calculated. All contributions are accumulated to obtain the light field encoded data of the pixel. The hologram calculation method proposed in this invention can guarantee the coherence between pixels, restore the true occlusion relationship of the model, and reduce the number of samples.

The application belongs to the technical field of automatic driving, and provides a positioning and map construction synchronization method, device, medium and vehicle. The positioning and map construction synchronization method comprises the following steps: acquiring a depth image of a surrounding environment of a vehicle; performing instance segmentation on the depth image of the surrounding environment of the vehicle, and simultaneously tracking a pose background area of an image acquisition device to determine each visible object model; tracking each visible object model and a geometric structure to jointly optimize the geometric shape and the pose of each visible object, and integrating overlap rate data of the joint optimization result and the instance segmentation to determine an initialized new object; based on the tracked initialized new object, the geometric shape and the pose of the initialized new object are jointly optimized again, overlap rate data of the re-joint optimization result and the instance segmentation are integrated again, and finally the object with the most accurate geometric shape and pose is integrated, and is positioned and displayed in an instantiated map synchronously.

The objective is to provide a display control device, a display control method, and a display control program that can improve the safety of persons present around invisible objects. The display control device (410) includes an object information acquisition unit (13) that acquires object information, which is information relating to an invisible object (T), and a display control unit (40) that generates an object information image (I1) showing the object information and controls a display device (210) using augmented reality technology so that the object information image (I1) is displayed in a way that is visible to a person (U) present in the vicinity of the invisible object (T).

A selector selects at least one piece of virtual object data from plural pieces of virtual object data based on first location information, first shape information, plural pieces of second shape information, and plural pieces of second location information. The first location information indicates a location of a user in a real space. The first shape information indicates a shape of a real object located in the real space. Each piece of second shape information indicates a shape of a corresponding one of the virtual objects. Each piece of second location information indicates a location in which a corresponding one of the virtual objects is arranged in the real space. A transmission controller transmits virtual object data to a display device. The selector estimates at least one visible object visually recognizable from the location of the user and selects virtual object data corresponding to the at least one visible object as the at least one piece of virtual object data.

The invention provides an invisible object 4D character interaction generation method based on text description, which comprises the following steps: stage 1, 3D person-object interaction key frame recovery: acquiring a human body motion sequence through a human body motion model and uniformly downsampling, and extracting a key frame; reconstructing a human body grid for each key frame through an SMPL-X model and extracting vertex positions to form a human body point cloud; the object position anchoring network predicts the object position by taking the human body point cloud, the object template point cloud and the text prompt as input, and generates a 3D human-object interaction key frame; stage 2, generation of a 4D human-object interaction sequence: constructing a contact perception diffusion model, taking a sparse 3D human-object interaction key frame as input, and extracting a condition signal through a contact perception encoder; and based on the condition signal, performing time sequence interpolation on the 3D human-object interaction key frame through a contact perception diffusion model, and generating a dense 4D human-object interaction sequence with a coherent time sequence. According to the invention, natural and vivid 4D character-object interactive synthesis of an unseen object is realized.

The invention provides a texture resource management method and device, a storage medium and electronic equipment, and relates to the field of computers. The electronic device determines an invisible object leaving the visual area in response to an update operation of a user on the visual area; a target texture corresponding to the invisible object in the video memory is determined, if no target visible object is left in the visible area, video memory resources occupied by the target texture are recycled, and the target visible object is a display object depending on the target texture in the visible area. In this way, the texture resources which are not quoted by the visual object are recognized and released in time, and therefore the risk of excessive consumption of the video memory is relieved.

The application discloses an information display method and device, a storage medium and electronic equipment, wherein the method comprises the following steps: displaying a map page, wherein the map page is a map page with a positioning information as a center point and a first scale; in response to a query operation, obtaining a query result, determining a second scale of the map page based on a display range corresponding to the query result, performing zoom processing on the map page based on the second scale, and marking a visible object corresponding to the query result in the zoomed map page. In the case that the user performs a query operation, the map page is adaptively zoomed according to the query result, and the visible object corresponding to the query result is marked on the zoomed map page, realizing automatic interaction between the user and the map page, avoiding the case that the user manually adjusts the map page to display the query result, and improving the flexibility of information query in the map page.

Systems and methods are provided for filtering and / or blocking unreliable content that might normally be injected in an Augmented Reality (AR) or Virtual Reality (VR) device. In one implementation, a method includes a step of analyzing a scene obtained by an image capture device associated with an Augmented Reality (AR) system to detect one or more visible objects in the scene. The method also includes using identifying characteristics of the one or more visible objects to obtain content related to the one or more visible objects. Before presenting the content in an augmented content overlay with respect to the AR system, the method further includes filtering the content based on one or more trust or validity factors.

A method for irradiating an image in an optical see-through (OST) head mounted display (HMD) for viewing through, the OST HMD by a user's eye, an object having at least one of known orientation and position and orientation (O / P&O), associated with a first reference frame, the method comprising: generating and irradiating said image for appearing to said user superimposed in an aligned manner to said object, according to predetermined information, eyeball feature position data, and said O / P&O; said predetermined information relates correction data with a plurality of different respective position data values of at least one eyeball feature position of said eye; said predetermined information further includes display corrections of said electro-optical display module with respect to said position data values of said at least one eyeball feature position, with respect to a second reference frame; and said O / P&O is between said second reference and said first reference frame.

An embodiment of the present application provides a labeling method for autonomous driving simulation, including: obtaining radar point clouds of the same type; obtaining instance numbers of each point in the radar point cloud of the same type, and generating a visible object number list based on the instance numbers of each point; generating a matching list based on the radar point cloud of the same type and the visible object number list, wherein the matching list includes multiple indexes and multiple contents, the multiple contents and the multiple indexes correspond one-to-one, at least one index includes the instance number in the visible object number list, and the multiple contents include points in the radar point cloud of the same type; obtaining camera information, the camera information includes the intrinsic parameters, extrinsic parameters and position information of the virtual camera that outputs the image to be labeled in three-dimensional space; and labeling the two-dimensional bounding box of the detected object on the camera plane of the virtual camera based on the matching list and the camera information.

The application discloses a learning object attention analysis method, a terminal and a device in an educational meta-universe, generates a visible object set corresponding to a learning object according to a virtual scene of an educational meta-universe space, and determines geometric-attribute information of each visible object in the visible object set; obtains a learning video of the learning object, determines a line-of-sight direction and a micro-expression of the learning object according to the learning video; determines a target visible object that the learning object gazes at according to the geometric-attribute information of each visible object and the line-of-sight direction, and associates the target visible object with the micro-expression; and analyzes the attention of the learning object according to the target visible object and the associated micro-expression, so that the attention of the learning object in the educational meta-universe can be comprehensively and accurately analyzed.