1537 results about "Virtual image" patented technology

The invention provides an interactive automatic explanation method for converting a traditional video into an artificial intelligence digital human, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining original video data and an audio track, and carrying out the semantic analysis of the audio track, and obtaining multi-mode deconstruction data; generating an explanation script for each time period of the video based on the explanation text, and performing timestamp labeling on the visual elements to form a time sequence synchronization data structure; in the playing process, a virtual image generator is driven to synthesize digital human dynamic expression output in real time according to the current playing time point; after a user interruption request is received, semantic matching is carried out on a query intention in the explanation script, a target explanation fragment and visual elements are positioned, and complementary explanation content is generated; and driving the virtual image generator to synthesize dynamic output synchronized with the supplementary explanation, and after interaction is completed, recovering playing or skipping to a specified time point according to a user instruction. According to the invention, the conversion from the traditional video to the interactive intelligent explanation video is realized, and the watching experience and learning efficiency of the user are improved.

The invention discloses a double-domain artifact correction method based on multi-energy-level data and physical prior fusion, and the method achieves the efficient and precise removal of dispersive metal artifacts through the construction of a complete technical scheme of the combination of multi-energy-level data collection, double-domain cooperative correction and physical model constraint. The system has the beneficial effects that the system covers multiple fields of medical treatment, industry, security and protection, aerospace and the like, and has extremely high universality and adaptability. On the basis of a multi-energy-level slow switching scanning protocol, full-angle scanning of at least two energy levels is completed by dynamically adjusting radiation source parameters, and obtained complete multi-energy-level projection data is converted into a high-dimensional tensor through a channel dimension splicing image fusion method. The constructed multi-channel virtual image completely retains attenuation characteristics and structure information of a target object under each energy, provides a more comprehensive input source with discrimination for a deep learning network, builds a data basis for accurate correction in different fields fundamentally, and adapts to various imaging scenes containing metal targets.

The invention relates to the technical field of vehicle vision calibration, and discloses a DMS camera calibration method based on computer vision. The method comprises the following steps: acquiring a multi-view original image sequence in a target vehicle cockpit, and synchronously acquiring a camera mounting position, a view angle, an inclination angle and an environment illumination intensity parameter; after distortion correction is carried out on the original image sequence, edge features of a calibration reference object in each frame of image are extracted through image recognition, and an initial calibration matrix containing a mapping relation between pixel coordinates and physical coordinates is generated; constructing a three-dimensional space projection model based on the initial calibration matrix, and adjusting rotation and translation parameters of the model by using an iterative optimization algorithm to enable the error between a virtual imaging position and an actual edge feature to be minimum; and extracting an internal parameter matrix and an external parameter matrix of the optimized model, fusing the internal parameter matrix and the external parameter matrix into a final calibration parameter, and writing the final calibration parameter into a DMS camera firmware storage area. According to the method, the DMS camera calibration accuracy and environmental adaptability are improved, and stable technical support is provided for driver state monitoring.

The invention discloses a video generation method, belongs to the technical field of artificial intelligence, particularly relates to the technical fields of computer vision, deep learning, large models and the like, and can be applied to scenes such as digital humans and the like. The method comprises the following steps: extracting a target audio feature of a target audio and a virtual image feature of a virtual image in a reference image; based on an attention mechanism, processing the target audio feature and the initial video feature to obtain a target video feature, the initial video feature being formed by splicing a virtual image feature and random noise; and decoding the target video feature to obtain a target video, the target video comprising a video frame of the virtual image expressed based on the target audio speech. The invention further provides a video generation device, an intelligent agent, electronic equipment and a storage medium.

The invention provides a video generation method and device, electronic equipment, a storage medium and a program product, relates to the technical field of artificial intelligence, in particular to the technical fields of computer vision, deep learning, large models and the like, and can be applied to scenes of digital human, content generation based on artificial intelligence and the like. The method comprises the following steps: acquiring text features of a description text, image features of a virtual image in a reference image and audio features of audio, wherein the description text indicates action description information for driving the virtual image based on the audio; the role feature is bound with the audio feature to obtain a target audio feature, the role feature is associated with a corresponding virtual image in the reference image, and the role feature is used for indicating an association relationship between the target audio feature and the image feature; and generating a target video based on the text feature, the image feature and the target audio feature, wherein the target video comprises a plurality of video frames for making a sound according to the action description information based on the audio driving virtual image.

The invention discloses an airborne visible light image automatic splicing method, and relates to the field of unmanned aerial vehicle image processing, and the method comprises the steps: obtaining a plurality of images collected in the flight process of an unmanned aerial vehicle, and the corresponding spatial position information and attitude information; on the basis of the spatial position relation between the images and the image overlapping information, determining image pairs capable of being registered, and constructing a connection map with the images as nodes and the image pairs capable of being registered as edges; detecting whether spatial connection fracture caused by image missing exists in the atlas or not, if so, inserting a virtual node and constructing a virtual image comprising an edge region and a transition region; and further calculating geometric transformation parameters between the images, resampling all image contents to a unified coordinate system, and executing pixel-level fusion processing in an image overlapping region. According to the invention, the problem of discontinuous image splicing of the unmanned aerial vehicle due to the existence of a no-fly zone, a privacy protection zone and the like of the unmanned aerial vehicle is solved.

The invention relates to the technical field of numerical control, in particular to an engraving and milling machine machining track control method based on machine vision. Comprising the steps of receiving an operation task instruction, collecting a reflected light intensity matrix and processing thermal field data, and generating a source data packet; obtaining a fusion heat distribution matrix through the homography matrix; based on the friction thermal response coefficient, a physical tool nose contact area is judged, a dynamic feature mask is generated, the dynamic feature mask and the reflected light intensity matrix are operated, and a monitoring area is output; executing end coordinates are output through least square ellipse fitting, and sensor calibration parameters are inversely transformed into tool nose feedback coordinates; and reading interpolation instruction coordinates, calculating a differential vector, and driving a feed shaft motor to execute displacement compensation. According to the method, deep space-time coupling is performed on the physical cut-off characteristic of the infrared spectrum, the high-frequency texture characteristic of the visible light and the kinematics constraint of a numerical control system, so that a real tool nose hot spot and an optical virtual image are distinguished, and the uniqueness and authenticity of track identification are ensured.

The invention discloses a vehicle-mounted head-up display system, a vehicle and a vehicle display driving method, and relates to the technical field of vehicle-mounted display.The vehicle-mounted head-up display system comprises a first display module and a second display module; the first display module is configured to project an image shot by the streaming media rearview mirror system out of a windshield in front of a driver to form virtual image display, and the second display module is configured to project head-up display content out of the windshield in front of the driver to form virtual image display. Therefore, the head-up display content and the image shot by the streaming media rearview mirror system form an integrated vision field, and a driver can simultaneously watch the front road, the head-up display content and the image of the streaming media rearview mirror system without shifting the sight line and the focal length, so that the driving safety is greatly improved; and a fused and coherent virtual image picture can be constructed, so that the driving experience is improved.

The invention relates to a virtual costume matching recommendation method and system based on a knowledge graph, and the method comprises the steps: obtaining virtual image data of a user, extracting image feature data of the user from the virtual image data, and constructing a dynamic user portrait in combination with the historical purchase record, social behavior data and preference setting of the user; constructing a costume knowledge graph, wherein the costume knowledge graph comprises costume attribute nodes and costume relation edges; based on a semantic matching algorithm of graph embedding, performing embedded representation on the dynamic user portrait and the clothing knowledge graph, calculating semantic similarity between the dynamic user portrait and the clothing knowledge graph, and generating a semantic matching result; based on a logical reasoning technology, according to scene requirements set by the user, reasoning to generate a clothing collocation combination of the user, and generating a logical reasoning result; and according to a semantic matching result and a logical reasoning result, generating a costume matching suggestion, and supporting a user to carry out interactive adjustment and feedback. The method and the device have the effect of improving the clothing recommendation accuracy.

The invention provides a VR visual training method and system based on dynamic refraction adjustment, and is applied to the technical field of VR visual training, and the system comprises a display module, a focusing mechanism, a hybrid optical system and a main controller. The main controller generates a target diopter sequence changing along with time according to the training mode, and drives the optical lens group to move along the optical axis direction, so that the focal power of the system is physically changed; meanwhile, the virtual image depth is updated according to the diopter, and perspective parameters are adjusted and rendered, so that the position of the presented virtual image is consistent with the optical focus in real time; in the training process, the adjusting reaction speed and the adjusting lag amount are obtained through the clear recognition performance of the user on the sighting mark, the main controller conducts self-adaptive adjustment on the follow-up refraction change track based on feedback, and training stimulation is made to be matched with the instant adjusting ability of the user. A real, continuous and individualized adjusting stimulation environment can be constructed, and the effectiveness and safety of ciliary muscle training are improved.

The invention relates to the technical field of smart city public safety, in particular to an urban gas pipeline leakage risk intelligent early warning and positioning method and device.Firstly, a digital twinborn construction and dynamic coupling module constructs and drives a virtual image of a gas pipe network, then, a multi-modal data acquisition and fusion module comprehensively senses the state of the pipe network, and finally, the leakage risk of the gas pipeline is monitored. The intelligent early warning and positioning decision module realizes a core intelligent algorithm; a cloud-edge-end collaborative architecture is adopted, and a terminal sensor is responsible for data acquisition; the edge computing node is deployed in a regional pressure regulating station and is used for carrying out primary data processing and local real-time alarming; the cloud center deploys a digital twin platform and a core AI algorithm model, and performs big data analysis, simulation deduction and global decision. Compared with the prior art, the method can realize early and accurate identification and prediction of gas leakage and potential risks (such as abnormal stress) of the pipeline.

The invention relates to the technical field of computer vision processing, in particular to an automatic defect positioning method based on a BIM virtual image. The method comprises the following steps: establishing a high-rise building model, and making a data set integrating an illumination condition and a full view angle; calculating camera parameters; performing cross-modal image retrieval: performing targeted optimization on the basis of a classical ResNet architecture, and constructing a backbone network structure suitable for a building image retrieval task; initializing position attitude estimation based on matching; and correcting the camera position posture. According to the method, a defect fixed frame based on a building BIM virtual image is provided, a building image data set BIM-Vision based on Revit is constructed, rich visual angles and illumination condition setting are achieved, accurate camera position postures and 3D labels between beam columns are provided, high-quality basic data support is provided for building visual research, and the method has the advantages of being high in practicability and high in practicability. And during inspection, accurate positioning of defect positions and component association can be completed only by shooting a field image, so that the field operation process is greatly simplified.

A method for controlling autonomous driving of a vehicle is introduced. The method may comprise, training, based on an inference depth and an inference pose, a synthetic image model for generating a synthetic image, generating, based on the synthetic image, a first virtual image to be associated with the original image, generating, based on the original image, a second virtual image, training a generative adversarial network (GAN) for determining, based on the original image, authenticity of the first virtual image and the second virtual image, training, based on the trained GAN, a depth network, wherein the trained GAN outputs a determination of the authenticity of the first virtual image, outputting, based on the trained depth network, signal, and controlling, based on the signal, autonomous driving of the vehicle.

The invention discloses an underwater multi-medium refraction parameter calibration method based on a binocular camera, and belongs to the technical field of underwater binocular vision measurement. The method comprises the following steps: firstly, completing Zhang Zhengyou calibration of a binocular camera in the air, and obtaining internal and external parameters and a distortion coefficient; establishing an'air-glass-air 'refraction model, and calculating a normalized incident ray direction by extracting checkerboard angular point pixel coordinates and performing distortion removal; constructing a position relation between a real target plane and a virtual image plane, and solving a normal vector of a refraction interface; combining a ray tracing method to construct a multi-dimensional constraint objective function fusing a physical spacing error, a plane residual error and ray intersection consistency, and adopting a nonlinear optimization method to jointly estimate the glass thickness and the distance from the optical center of the camera to a refraction interface; and finally, expanding to an air-glass-water model to realize high-precision three-dimensional positioning of an underwater target point. The method does not need underwater calibration, and is suitable for scenes of underwater structure detection, submarine topography modeling and the like.

Systems and methods of virtual displays to generate virtual images. Some display systems comprise a light source and a plurality of specular reflectors to produce virtual images that have a monocular depth that is closer to a viewer than an exit aperture optic or other optical component of the display system. In some embodiments, sensors or cameras detect a property of a viewer or of an environment to impact the virtual image or to provide sensory feedback to the viewer.

An optical system is presented for use in a near-eye mixed reality system. The system comprises a relay system defining an eyebox of the optical system, said relay system being configured and operable to relay a virtual image light field from a light-engine onto an eye pupil plane while combining said virtual image with real-world light field. The relay system is configured as a free space relay system configured for free space propagation of said virtual image light field being relayed, said free space relay system comprising at least one off-axis 4f-system. Each of said at least one off-axis 4f-system comprises at least one lens formed from at least one resonance-domain surface relief diffractive optical element (SRDOE) operable for combining said virtual image light field with the real-world light field, said at least one SRDOE being configured with a predetermined global surface relief pattern characterized by global variation of at least some of pattern parameters across said SRDOE.

A holographic near-eye display device with multi-angle simultaneous illumination comprises a light source module, a spatial light modulator, a beam splitter, an eyepiece, and a master controller. The light source module emits parallel light at different angles and simultaneously illuminates and covers effective working area of spatial light modulator which, loaded with a hologram, modulates incident parallel light at different angles to form diffracted parallel light at different angles, that is, virtual images at different viewing angles. Parallel light at different angles illuminates spatial light modulator. After modulated and diffracted by calculated hologram on spatial light modulator, the diffracted image light is converged by second lens to form different viewpoints for human eye to view. In multi-angle simultaneous illumination, additional time-sharing control of illumination unit is not needed. When size and position of human eye pupil change, clear virtual image is always seen, achieving expanding the eyebox.

The application discloses a rotating mirror type binocular vision laser tracking light break recovery system and method. The system comprises a rotating mirror type laser tracker and a binocular vision module, two cameras in the binocular vision module are symmetrically installed on both sides of a laser beam, and camera visual axes and a laser optical axis are reflected to a working area through a rotating mirror of the tracker. The method comprises the following steps: a virtual imaging plane model in which the camera visual axes and the laser optical axis are coincident is constructed in advance; after the light break occurs, an infrared illumination light source of a vision system is modulated, frame difference images of modulation wave crests and troughs are acquired and processed, a current position of a target mirror and a predicted light break recovery rotation angle are obtained, and error compensation is performed on the current position and the predicted light break recovery rotation angle; and corresponding motors are driven to rotate according to the compensated light break recovery rotation angle, so that the laser beam is re-aligned to the target mirror. The application reduces the coaxial installation difficulty of a vision system of a rotating mirror type laser tracker, and improves the target mirror recognition rate and the light break recovery servo angle accuracy in a large range of complex environments.

The embodiment of the invention provides an augmented reality board game system and a control method, and relates to the technical field of augmented reality. The system comprises at least two interaction subsystems in communication connection, each interaction subsystem comprises a physical game device, an image acquisition device, a data processing unit and an augmented reality display device, the physical game device comprises a datum reference structure and a plurality of physical game components, the image acquisition device is used for acquiring image data of the physical game device, and the data processing unit is used for processing the image data. The data processing unit is used for calculating coordinate information of the entity game component in a three-dimensional world coordinate system based on the image data, and the augmented reality display device is used for rendering a virtual image of the entity game device according to the coordinate information of the game components of the local interaction subsystem and other interaction subsystems in the three-dimensional world coordinate system. Therefore, a plurality of different-place players can be supported to operate the entity game component through real touch control, and share the same augmented reality fused board game scene in real time.

Aspects of the method, apparatus, non-transitory computer readable medium, and system include estimating extrinsic camera parameters of a plurality of cameras based on detection of corners in a camera calibration arrangement by projecting an image of the detected corners onto an image plane of at least one of the plurality of cameras. The aspects further include incrementally tuning a location estimate for the detected corners based on pixels of the projected image, and adjusting the extrinsic camera parameters of the at least one of the plurality of cameras based on the tuned location estimate. The aspects further include generating a virtual image based on the pixels of the projected image mapped to a virtual image plane.

The invention discloses a naked-eye 3D augmented reality interactive display system, and relates to the technical field of naked-eye 3D display and reality interaction. According to the invention, the multi-source depth data acquisition module acquires and fuses a user limb depth image, the parallax physical depth mapping module completes depth and parallax mapping and visual angle correction, the occlusion relation determination module divides regions to determine the occlusion type, and the adaptive mask layer generation module generates a dynamic smooth mask layer. The shielding effect fusion display module is used for overlapping and synthesizing the shielding layer and the multi-view virtual image and then displaying the overlapped and synthesized shielding layer and the multi-view virtual image; the problems that in naked eye 3D interaction, the occlusion relation between the limbs of the user and the virtual object is not accurate, and the vision is abrupt are effectively solved, the space consistency and the vision fluency of interaction are remarkably improved, and the real and natural naked eye 3D augmented reality interaction experience is achieved.

The invention discloses a near-ground distance measurement method based on virtual large baseline four-eye vision, a medium and equipment, and the method comprises the steps: synchronously obtaining four images when a lifting appliance enters a near-ground operation range; based on the calibration relation between the cameras, the high-altitude images are respectively projected and transformed to the visual angles of the corresponding lifting appliance cameras and fused, two virtual camera images are generated, and therefore a virtual large base line far exceeding the physical distance is constructed; performing three-dimensional correction and cutting on the virtual image pair to obtain a row-aligned three-dimensional image pair; performing feature matching and triangulation on the virtual image pair to generate a sparse reference depth map; and inputting the stereo image pair and the sparse depth map into a pre-trained depth estimation neural network model together, outputting a dense depth map, and converting the dense depth map into a coordinate system taking the lifting appliance as an original point to obtain a vertical distance. According to the invention, the ultra-large baseline is virtually synthesized by using the existing camera, the remote distance measurement precision is improved, the cost is low, and the reliability is high.

The invention discloses a virtual image clothes changing method and system, and the method comprises the steps: carrying out the preprocessing of a figure image inputted by a user, and obtaining standardized image data; recognizing a face region in the standardized image data, and constructing a three-dimensional face model; body parameters corresponding to the three-dimensional face model are determined, and a three-dimensional human body model is generated based on the body parameters; based on a matching result between the measured size of each key part in the three-dimensional human body model and a design size of a corresponding part in a preset clothing template, deforming the clothing template to obtain a three-dimensional clothing model, and simulating virtual clothing changing effect data when the three-dimensional human body model wears the three-dimensional clothing model; and updating the three-dimensional human body model according to the facial features of the user and the body posture of the user, updating the virtual clothes changing effect data according to the updated three-dimensional human body model, and obtaining a virtual image preview picture corresponding to the dynamic performance behavior of the user.

The invention discloses a multi-focal near-to-eye display holographic optical element and a preparation method and device thereof. The element adopts a composite structure of a substrate, a multi-focal correction layer and a holographic recording layer, a holographic grating accurately matched with each optical partition of the correction layer is formed in the recording layer through a partition exposure technology, and multi-depth virtual image display and user refraction correction functions can be realized at the same time. The preparation device comprises a coherent light source, a beam splitter, a diffusion plate, a focal plane displacement system, a mask component and a master controller, and partition exposure is realized through cooperative control. The problem that traditional near-to-eye display cannot meet multi-focus display and refraction correction at the same time is solved, a user can clearly perceive fused multi-depth virtual information and a real environment without additionally wearing glasses, the visual comfort and experience are remarkably improved, and the method is suitable for AR / VR and other head-mounted display devices.

The invention discloses a multi-modal driving system based on streaming big language model output, which belongs to the technical field of multi-modal interaction and comprises the following steps: a streaming big language model module receives input information and then generates a text unit sequence in a streaming manner; when each text unit is generated, a multi-dimensional emotion label sequence containing emotion types, emotion intensity values and semantic association degrees is synchronously output; the speech generation module maps the text unit sequence and the corresponding multi-dimensional emotion label sequence into speech synthesis parameters, and drives a streaming text-to-speech model to generate an emotional speech stream; the label generation module generates an expression label sequence and an action label sequence of the virtual image in real time; the emotion synchronous control module performs synchronous interpolation and coordination control by taking the time step of the text unit as a reference to generate a synchronous multi-mode driving signal; and the rendering and driving module drives the virtual image based on the signal to realize collaborative output of voice, mouth shape, expression and limb movement, and realizes synchronous presentation of multi-modal emotion expression.

The invention relates to an improved cage type light field imaging system and a debugging method, and belongs to the technical field of light field imaging, and the system comprises a main lens which is used for collecting light spots on an object surface, and enabling the light emitted by the light spots to be imaged on a micro lens; the micro lens is used for separating the light collected by the main lens into light beams in different directions and imaging the light beams on a virtual image plane; the relay lens is used for carrying out secondary imaging on the light beams in different directions to different pixels of a CCD detector of the camera; the relay lens comprises a first lens and a second lens, the focal lengths of the first lens and the second lens are positive, the focal length of the first lens is smaller than that of the second lens, and the magnification of the relay lens is larger than 1: 1. In the debugging process of the system, parallel light beams are arranged in front of the micro lens, and the position of the micro lens is determined in the mode that the parallel light beams vertically enter the micro lens. The system is not influenced by vignetting and distortion generated by the relay lens any more, and the debugging method improves the positioning precision of the micro lens.

The utility model discloses an off-axis bifocal plane HUD optical system, which comprises an image source, an eye box, a lens, a first reflecting mirror, a second reflecting mirror and an automobile windshield, the image source comprises a first output part and a second output part, light of the first output part passes through the first reflecting mirror, the second reflecting mirror and the automobile windshield to enable the eye box to obtain a near focal plane, and light of the second output part passes through the first reflecting mirror, the second reflecting mirror and the automobile windshield. The light of the second output part passes through the lens, the first reflecting mirror, the second reflecting mirror and the automobile windshield to enable the eye box to obtain a far focal plane. The image source is divided into two parts, the lens is arranged on the light path of the second output part of the image source, and a near-focus virtual image and a far-focus virtual image are formed in front of the windshield, so that a driver can see the virtual image surface and timely obtain the driving information of the vehicle under the condition of poor visual field such as vehicle following or traffic jam, and the driving safety of the vehicle is improved. Therefore, the thickness of the crystalline lens can be adjusted when human eyes watch far and near focuses, and fatigue caused by long-time observation of a certain focal plane is avoided.

The invention provides a head-up display system and a vehicle, and belongs to the technical field of optics. The head-up display system comprises an image source, a first plane lens, a second plane lens and a free-form mirror. The image source includes a first display area and a second display area. The first plane lens is located on an emergent light path of the first display area, and the second plane lens is located on an emergent light path of the second display area. The optical path of light emitted by the first display area on the first plane lens is different from the optical path of light emitted by the second display area on the second plane lens. The free-form mirror is located on transmission light paths of the first plane lens and the second plane lens. Because the optical path of the light in the first plane lens is different from the optical path of the light in the second plane lens, the equivalent object distances of the first display area and the second display area are different, and the virtual image distances of two focal planes corresponding to the first display area and the second display area are different. Therefore, the head-up display system can generate two focal planes with different virtual image distances.

The invention relates to the technical field of telemedicine, and particularly discloses an online inquiry method and device, computer equipment and a storage medium. According to the method, the doctor can be accurately matched according to the inquiry information and the historical medical data of the patient, the inquiry information and the historical medical data are analyzed, the to-be-examined part and the examination prompt of the patient are determined, effective reference is provided for diagnosis and treatment, manual analysis is not needed, the virtual consulting room is constructed through the augmented reality equipment, and the diagnosis efficiency is improved. Patient feedback can be obtained in time according to the patient virtual image, and the diagnosis efficiency is improved. The method is applied to an online inquiry system in the medical field, inquiry information and historical medical data of a patient can be accurately analyzed, and a doctor is effectively assisted in quickly and accurately obtaining a diagnosis result.

Disclosed herein are computer-implemented method, system, and computer-readable medium embodiments for providing a virtual second image during remote deposit of documents. A system generates, from a field of view of a camera, a live stream of image data comprising imagery of a portion of the document. The system obtains, via an optical character recognition program, one or more data elements from the imagery of the portion of the document and generates a first image of a front side of the document. The system receives an electronic signature of a customer via a user interface and, upon verifying the electronic signature, generates a virtual second image by merging the electronic signature with a second image for a back side of the document selected from a plurality of back images. The system transmits the first image, the virtual second image, and the one or more data elements to a remote deposit platform.