46 results about "Path tracing" patented technology

Approaches presented herein provide systems and methods for path tracing using a set of textured spherical surfaces obtained from an importance map for an image. An image representation may be generated using the importance map and evaluated to identify a first set of nodes. The nodes may have associated values, such as luminance values, that may be used to subdivide the nodes into bins to maintain a weighed distribution for the associated values. An array of nodes may be generated for sampling and conversion to a three-dimensional direction that may be applied to one or more lighting effects.

Disclosed approaches provide for interactions of secondary rays of light transport paths in a virtual environment to share lighting contributions when determining lighting conditions for a light transport path. Interactions may be shared based on similarities in characteristics (e.g., hit locations), which may define a region in which interactions may share lighting condition data. The region may correspond to a texel of a texture map and lighting contribution data for interactions may be accumulated to the texel spatially and / or temporally, then used to compute composite lighting contribution data that estimates radiance at an interaction. Approaches are also provided for reprojecting lighting contributions of interactions to pixels to share lighting contribution data from secondary bounces of light transport paths while avoiding potential over blurring.

The invention discloses an automatic driving laser radar simulation method for high-fidelity material modeling, and the method comprises the steps: carrying out the high-fidelity modeling of a complete physical process of a single laser channel, and designing a complete technical chain from laser beam modeling, material reflectivity modeling, light path modeling to post-processing. In laser beam modeling, energy space distribution is accurately depicted by introducing an angle attenuation conical beam model; in material reflectivity modeling, a bidirectional reflection distribution function system covering Lambert body, specular reflection, transmission, retroreflection and general materials is established; in light path modeling, traditional single reflection hypothesis is broken through, time-resolved Monte Carlo path tracking and kernel density estimation are adopted, simulation of multi-path reflection and transmission effects is achieved, traditional model bottlenecks are broken through, complex optical phenomena are simulated, the generation mode has expandability and controllability, rare key scenes are efficiently synthesized, and the method is suitable for large-scale popularization and application. And the method has wide field extension potential and high technical universality.

The invention discloses an unmanned aerial vehicle image-assisted path tracking and obstacle avoidance method and system, and the method comprises the following steps: calling a multi-modal sensor group deployed on an unmanned aerial vehicle to collect environment data in response to a flight task instruction of the unmanned aerial vehicle, and carrying out the cross-modal data fusion of the environment data through an adaptive time synchronization algorithm; and calling a preset generative adversarial network model, performing image generation and illumination compensation on low-light image data in the fused environment data, and generating a target enhanced image close to a visible light environment. According to the invention, through cross-modal learning and self-supervised learning technologies, the perception and data synchronization capability of the unmanned aerial vehicle in a complex environment is improved. According to the invention, the generative adversarial network is adopted for image enhancement, and especially in a low-light environment, the accuracy of obstacle detection is improved; real-time obstacle detection and three-dimensional environment modeling are performed in combination with a deep learning model, and path planning and obstacle avoidance control are optimized through deep reinforcement learning.

The invention discloses a product oil supervision beforehand early warning system and method based on AI and big data fusion, and relates to the technical field of data supervision. Comprising the steps of 1, constructing a real-time monitoring network, deploying a millimeter-wave radar liquid level instrument to track the liquid level change of an oil tank in real time, matching a refueling machine tax control sensor and an AI camera for behavior recognition, monitoring refueling operation in a whole process, and collecting whole-process data, 2, integrating heterogeneous data sources, carrying out cross-department data real-time synchronization through Kafka and DataX, and carrying out real-time monitoring on the whole-process data. The method comprises the following steps: step 1, establishing a supervision data lake, and preprocessing the collected data to form a structured data system covering the whole chain of purchase-sales-storage-transportation, and step 2, fusing a multi-modal AI model to identify non-periodic abnormity of the liquid level and temperature of an oil tank and early warning equipment associated faults according to the preprocessed structured data, the method comprises the following steps: step 1, constructing an oil station three-dimensional model, fusing transaction data, monitoring videos and equipment logs, identifying cheating behaviors, and constructing an oil station-equipment-personnel-transportation knowledge graph for path traceability of quality abnormity and cheating risks, and step 4, constructing a 1: 1 oil station three-dimensional model by adopting laser point cloud scanning, and automatically highlighting a fault area and performing whole-process visual display when early warning is triggered.

The present application relates to the field of photoelectric detection technology, and provides a full-airspace target capture and tracking method based on a five-axis decoupling control system, comprising: obtaining the initial position parameters of the target through a five-axis decoupling control system; wherein the five-axis decoupling control system comprises: a two-degree-of-freedom swing mirror, a first fixed reflector, a second fixed reflector, a fast reflector and a photoelectric detector; constructing a multivariate linear equation group based on the final position parameters determined by the photoelectric detector and the initial position parameters of the target; using the original-dual path tracking method with predictive correction to solve the multivariate linear equation group in real time, obtaining the optimal angle adjustment amount of each axis, and giving the solution result; a servo control unit controls the angle adjustment amount of each axis in the five-axis decoupling control system according to the solution result, so as to accurately capture and track the target.

The invention discloses an automatic distribution method for patrolling and dispatching personnel based on a knowledge graph. The method comprises the following steps: S1, collecting and preprocessing data information; s2, constructing a heterogeneous knowledge graph, setting node types and generating an attribute matrix and an adjacent matrix; s3, traversing a multi-hop path, calculating an edge weight, and extracting a person post semantic feature vector; s4, introducing avoidance, duty rotation and conflict rules, and performing constraint screening on the semantic feature vectors; s5, scoring the candidate persons based on the post matching degree, the historical frequency and the post rotation period, and outputting a post matching score set; s6, sorting the post matching score set, and generating a sorting list according to the fitness from high to low; and S7, performing path traceability on the primary selection result, extracting associated information, and outputting a recommended result set of the calling personnel. According to the invention, the intelligent, compliant and interpretable distribution of the dispatching process of the patrol personnel is realized, and the personnel selection efficiency and the post matching accuracy are remarkably improved.

This invention relates to the field of unmanned surface vessel (USV) path tracking control technology, specifically disclosing a method, computer equipment, and storage medium for path tracking of USVs based on prioritizing experience replay optimization of near-end policy. The method includes setting up a prioritizing experience replay memory pool, an action space, a state space, and a reward function; constructing a network model based on a near-end policy optimization algorithm, and training the network model using samples from the prioritizing experience replay memory pool; generating current state information based on images acquired by the USV, the current position information of the USV, and the position information of matching points on the reference trajectory; and prioritizing the training of samples with poor action performance based on priority sampling data. This invention employs a prioritizing experience replay memory pool to ensure rapid model convergence, and the network model is constructed based on near-end policy optimization. The use of a state space and reward function tailored to environmental conditions improves the algorithm's adaptability and stability to the environment.

This application provides an image distortion simulation method and electronic device. The method includes: acquiring first image parameters of a source image, second image parameters of a target image, and lens parameters of a camera, wherein the target image is a distorted image of the source image after being imaged by the camera's lens; determining the spatial coordinates of each target pixel in the target image based on the second image parameters; performing reverse ray path tracing on each target pixel based on the spatial coordinates, the first image parameters, and the lens parameters to obtain the mapping relationship between each target pixel and each non-distorted pixel in the source image; assigning color values ​​to each target pixel in the target image based on the mapping relationship, and outputting the assigned target image to improve the accuracy, efficiency, and flexibility of image distortion simulation.

The invention discloses a pedestrian re-identification method based on path tracking, and the method comprises the steps: constructing a geographic time-space relation graph according to the layout position of a camera group, taking the camera group in which a target pedestrian appears for the first time as the camera group of the current round, and adding the camera group into a path record set; taking the current round and the adjacent camera group as a query camera group and a gallery camera group respectively; a monitoring video shot by a camera group is input into an image preprocessing module to extract pedestrian images to construct a query set and an image library set, and then the pedestrian images are input into a pedestrian re-recognition module to be compared to obtain a hit image; if the target pedestrian exists in the hit image, adding the corresponding camera into the path record set, taking the corresponding camera as a camera group of the current round, and repeating the previous steps; otherwise, outputting the target pedestrian path trajectory. According to the method, the application technical problems of huge detection quantity, low robustness of the pre-training model and poor applicability are solved, and the problem of intuitively describing personnel path tracking logic is solved.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that provide guidelines for path tracing of raster images. In particular, the disclosed systems, in one or more implementations, determine a set of outlines corresponding to the boundaries of a set of segments within a raster image. The disclosed systems select from the set of outlines an outline corresponding to a segment in response to input from a client device specifying one or more points within a threshold for the outline's distance. The disclosed systems provide a highlighted indication of the outline corresponding to the segment for display in a graphical user interface of a device.The disclosed systems generate a vector path within a vector image based on the outline corresponding to the segment, in response to a selection of the outline via the graphical user interface.

A method for rendering an image of a three-dimensional scene using path tracing. For a current pixel of the image to be rendered using path tracing, an associated pixel is identified, associated with the current pixel, wherein at least one parameter for rendering the associated pixel using path tracing has been previously determined. The at least one parameter for rendering the associated pixel is obtained, and used to determine at least one parameter for rendering the current pixel. The at least one determined parameter is output to control a rendering of the current pixel using path tracing.

This invention relates to a system and method for opinion extraction and evolution path tracking, belonging to the field of film review analysis technology. It includes a data input module, a data processing module, a temporal alignment module, a temporal enhancement module, a deep extraction module, an analysis module, and a visualization module. The data input module acquires long film review text data, temporal metadata, and external event data. The text processing module outputs text feature vectors, temporal decay factors, and external event feature vectors. The temporal alignment module outputs aligned temporal context. The temporal enhancement module performs temporal enhancement attention calculation. The deep extraction module extracts deep temporal features. The analysis module outputs sentiment curves and topic drift detection results. The visualization module generates a dynamic reputation evolution map and a visualization analysis report. This application effectively demonstrates the evolution of opinions in long film reviews.

The present disclosure relates to systems and methods for volume data rendering. The methods may include obtaining an initial image relating to volume data of a subject by performing a path tracing operation on the volume data, obtaining a total sampling operation count corresponding to the initial image, determining, based on the volume data, one or more reference parameters relating to the initial image, and generating, based on historical image data relating to the volume date of the subject, the initial image, the total sampling operation count, and the one or more reference parameters, a target image using an image processing model.

A method for rendering an image of a three-dimensional scene using path tracing. Scene feature data for a pixel of the image to be rendered using path tracing is obtained, indicating visual features of a location of the three-dimensional scene for depiction by the pixel in the image, is obtained. The scene feature data is processed at an artificial neural network, ANN, to determine a budget allocation parameter for rendering the pixel using path tracing. The budget allocation parameter is indicative of an amount of computing resources to be used for rendering the pixel using path tracing. The ANN is trained to determine, from the scene feature data, the budget allocation parameter based on the visual features indicated by the scene feature data. The determined budget allocation parameter is output from the ANN to control a rendering of the pixel using path tracing.

Systems and methods for 3D image scanners for real-time dynamic 3D surface imaging are disclosed. Embodiments of the present systems and methods describe a system and method comprising a first and / or second camera, and a projector, and a processor, the projector projects structured light with a fringe pattern onto a 3D object, the processor is configured to extract a phase map and a texture image from the image, and the processor calculates depth information from the phase map. Embodiments further describe methods and systems to determine wrapped phase from the image using a Hilbert transform, and to generate absolute phase from the wrapped phase using a combination of a quality-guided path following algorithm, a two-wavelength phase unwrapping algorithm, or a Markov Random Field method, and to generate a phase map from the absolute phase to determine depth information of the 3D object. Algorithms using conformal mapping, optimal transport mapping, and Tschupmueller mapping are used to register and track captured 3D geometry surfaces.

The invention discloses a calibration method of a curved surface medium imaging vision measurement system based on a light model. The calibration method comprises the following steps: shooting a group of first plane targets under the condition of no curved surface medium; extracting a first target feature, and performing camera calibration of the general distortion model based on the first target feature; shooting a group of calibration point images in first spatial distribution under the condition of no curved surface medium, and calibrating a first pose of the camera in a world coordinate system based on the calibration point images in the first spatial distribution; shooting a group of second plane targets in the presence of the curved medium; extracting a second target feature, and performing camera calibration of the general distortion model based on the second target feature; shooting a group of calibration point images in second spatial distribution in the presence of the curved medium, and calibrating a second pose of the camera in the world coordinate system based on the calibration point images in the second spatial distribution; eliminating distortion of the extracted second target feature and the second spatial distribution calibration point image feature by using a general distortion model calibrated based on the first target feature to obtain a distortion-removed two-dimensional feature point set; establishing a light model between an image point which is imaged through the curved surface medium and incident light outside the curved surface medium based on light path tracing; and calibrating mapping parameters of the light model by using the two-dimensional feature point set after distortion removal and the corresponding three-dimensional space position. A corresponding measurement method is also disclosed.

The application provides a violent and terrorist video recognition method based on key feature fusion. A key feature database based on violent and terrorist video recognition is built in advance; a to-be-recognized segment video is acquired, and content features and propagation channel data of the to-be-recognized segment video are extracted; wherein the content features include audio features and image features; according to the key feature database, audio key features and image key features of the to-be-recognized segment video are subjected to time-axis synchronous space-time fusion, and synchronous space-time features are generated; according to the propagation channel data, propagation path tracing and synchronous space-time feature authentication of the to-be-recognized video are performed, and a video subject of the to-be-recognized video is determined; according to the video subject, violent and terrorist video identification is performed, and when the video subject is a violent and terrorist video, the to-be-recognized segment video is identified as a violent and terrorist video.

The application discloses an infrared scene real-time light tracing rendering method based on a GPU hardware light tracing pipeline and belongs to the field of infrared simulation of virtual scenes. The application is based on the path tracing principle and takes Unity as a platform to realize real-time rendering of infrared scenes. Based on the infrared physics principle, firstly, real-time rendering of the infrared scene is realized based on the path tracing; on the basis, the GPU hardware light tracing pipeline technology is applied to improve the speed of the infrared path tracing, and the importance sampling and the space-time filtering denoising algorithm are used to guarantee the image quality of the infrared scene, so that the real-time rendering of the infrared scene is realized, the realism and details of the infrared rendering result are greatly improved, and the similarity is higher compared with the actually shot infrared scene image.

The invention discloses an infrared scene rendering method based on rasterization and path tracking hybrid architecture, and belongs to the field of image processing, and the method comprises the steps: building a corresponding heat transport tracking path for each heat transport mechanism in a virtual space; according to all the heat transport tracking paths, determining a comprehensive heat transport environment of a virtual space formed by the combined action of all heat transport mechanisms; rasterization rendering is carried out based on the comprehensive heat transport environment, so that an infrared image of the virtual space under the target virtual camera shooting view angle is obtained; and according to the infrared image, generating an imaging image under the target virtual camera view angle. According to the method, the authenticity of the infrared image is improved, the calculation cost and the visual angle sensitivity are reduced, and the adaptability of the infrared rendering process in a complex scene is enhanced.

Methods are provided for visualizing end-to-end path traversed by a hybrid media traffic flow. Specifically, the methods involve obtaining at least one hybrid flow record that includes information about a source of a traffic flow prior to performing a network address translation for the traffic flow. The traffic flow traverses from the source, along a path in a network, to a destination. The method further involves generating a merged flow data record in which at least two flow data records are merged based on the at least one hybrid flow record. The at least two flow data records include a unicast flow data record and a multicast flow data record based on the traffic flow traversing along the path in the network. The method further includes providing an end-to-end visibility of the path traversed by the traffic flow based on the merged flow data record.

The application provides a filter-based Monte Carlo gradient path tracking rendering image reconstruction method and device, wherein the method comprises the following steps: obtaining an initial rendering image obtained by Monte Carlo gradient path tracking rendering, and auxiliary information map in the rendering process; solving an optimized gradient by a gradient optimization objective function according to the initial rendering image and the auxiliary information map, and generating a multi-resolution level; performing filter calculation on the initial rendering image, the optimized gradient and the auxiliary information map according to the multi-resolution level from the lowest resolution level to the highest resolution level to obtain a target filter result; wherein the target filter result is a filter result output by the highest resolution level, and the filter result output by the highest resolution level is a reconstruction result of the initial rendering image. The method fully utilizes the gradient domain information and the auxiliary information map in the rendering process, solves the filter kernel coefficient at a resolution level, generates a high-quality and consistent denoising image (reconstruction result) at a low sampling rate, reduces the influence of noise in the gradient estimation on the reconstruction result, has a series of good characteristics such as consistency (gradual convergence to a reference image), and meets the efficient and high-quality rendering image reconstruction requirements.

The invention relates to the technical field of image processing, and discloses a cooperative control method and system for sectional dismantling and hoisting of a net rack in a complex environment. The method comprises the following steps: collecting net rack data through a multi-source sensor, and performing fusion processing to form a digital model; performing mechanical analysis and segmentation according to the digital model to obtain a segmentation scheme; connecting nodes are recognized according to the segmentation scheme, the dismantling sequence is optimized, and a dismantling path is formed; tracking segmented positions, and planning a hoisting track; robot tasks are coordinated, the working environment is monitored, and a cooperative control instruction is generated; and analyzing the execution data, recording deviation, and forming optimized data. According to the method, collaborative operation of multiple robots is achieved, an operation evaluation and optimization mechanism is established, a comprehensive integrated net rack segmented dismantling and hoisting collaborative control method is provided, and it is ensured that dismantling operation of a net rack structure can be safely and efficiently completed under the complex environment condition.

This invention discloses a method and system for optimizing dynamic rendering of virtual scenes based on light field reconstruction, specifically relating to the field of image data processing technology in computational imaging. First, it completes the access of multi-source data, geometric / optical calibration, and affine alignment of timestamps, constructing a unified time base and evidence chain within a sliding window of fixed step size. Then, it establishes a multi-scale spatial parameterized structure within the scene bounding box, using path tracing as the forward model to jointly solve geometry, radiation, and material properties, completing static / dynamic decomposition and fusing color, depth, polarization, and spectral residuals to generate an optical path consistency map. Based on this, it constructs a priority scalar field fused from uncertainty and importance, performing area-level budget arbitration and degradation control, and mapping it to executable control quantities such as sampling density, bounce depth, geometric level, and indirect light order. Finally, it improves cross-frame stability through temporal robust fusion and differential updates, and feeds quality metrics and operational statistics back to the next frame's scheduling to form a closed loop.

The approaches presented here provide systems and methods for path tracing using a set of textured spherical surfaces derived from an importance map of an image. An image representation can be generated and evaluated using the importance map to identify an initial set of nodes. The nodes can have associated values, such as luminance values, which can be used to subdivide the nodes into bins to maintain a weighted distribution for the associated values. An array of nodes can be generated for sampling and conversion into a three-dimensional direction that can be applied to one or more lighting effects.

The application relates to the technical field of image processing, and particularly discloses a rendering method and device based on photon mapping and path tracking, a computer device and a storage medium. First, photons are distributed according to light source information and object information, global light is simulated quickly by using photon mapping, a photon mapping diagram is obtained, unnecessary photon tracking is reduced, and the efficiency of obtaining the photon mapping diagram is improved. Then, the initial value of a light ray is determined according to the photon mapping diagram, path tracking is performed on the light ray emitted based on the initial value of the light ray, the propagation and reflection of the light ray can be effectively simulated, and a target rendering image is obtained. The rendering is realized by combining photon mapping and path tracking, the rendering efficiency is ensured, and the rendering quality is improved.

Methods, systems, and machine learning models for denoising deep images are disclosed. Deep image renders can be generated using Monte Carlo rendering methods such as path tracing. Unfortunately, such renders may have visual noise. A machine learning model according to embodiments can be trained to denoise deep images, and can comprise an embedding sub-model and a denoising sub-model. A computer system can use the embedding sub-model to generate a deep image embedding based on a noisy deep image input using novel local attention mechanisms. The computer system can use the denoising sub-model to denoise the noisy deep image using the deep image embedding. In some embodiments, the computer system can use the denoising sub-model to generate denoised deep images at multiple levels (or “scales”) and combine the denoised deep images to produce an output denoised deep image.

The invention provides an intelligent wind power blade inspection method and system based on image recognition, and relates to the technical field of wind power equipment inspection. Firstly, a wind power blade surface image set including continuous waypoint space position identification and continuous time frame sequence image data acquired by unmanned aerial vehicle inspection is acquired; and performing blade surface vibration response mapping processing on the continuous time frame sequence image data to obtain a blade surface vibration response mapping graph, performing vibration wave propagation path tracking processing on the mapping graph corresponding to the continuous spatial position identifier to generate a propagation path tracking result, and performing structure abnormality positioning processing according to the propagation path tracking result to obtain a blade surface vibration response positioning result. And generating a structure anomaly positioning result containing the vibration wave propagation path anomaly deflection point and the vibration wave energy attenuation anomaly region, and finally generating a wind power blade inspection report containing the spatial position label based on the structure anomaly positioning result. The structure abnormity of the wind power blade can be comprehensively and accurately detected, and the inspection efficiency and accuracy are improved.

Techniques are disclosed relating to ray tracing, particularly for path tracking through a bounding volume hierarchy, according to some embodiments. In some embodiments, ray intersection accelerator circuitry traverses a bounding volume hierarchy that includes transform nodes at multiple levels. Ray transform accelerator circuitry may transform ray information based on reaching a given transform node during the traversal. Path tracking hardware may determine a list of pointers to transform nodes in a traversal path taken to reach a first primitive. This may include utilizing at least two buffer circuits to store pointers for a candidate list and a committed list, storing candidate trail information that indicates, for a given position in the candidate list, which buffer stores a corresponding pointer, and storing committed trail information that indicates, for a given position in the committed list, which buffer stores a corresponding pointer.