63 results about "Visual capture" patented technology

The invention relates to the technical field of video monitoring, and discloses a camera cooperative monitoring method and system based on multi-modal perception, and the method comprises the steps: detecting an abnormal event signal in an environment through a non-visual sensor node, triggering the activation of a camera, and carrying out the visual capture of a target region; the camera performs relay tracking based on target feature matching to generate a continuous motion track; performing multi-view collaborative shielding on privacy sensitive information in the tracking target, generating desensitized monitoring data, and uploading the desensitized monitoring data; and optimizing the energy consumption of the camera in a non-event triggering period, and updating the behavior recognition model based on the desensitization data. According to the method, the contradiction between privacy protection and monitoring efficiency is solved, cross-regional model evolution and energy consumption reduction are realized, and the problems of response lag, data redundancy, privacy disclosure and high energy consumption of traditional monitoring are systematically avoided.

The invention relates to the technical field of parcel sorting, in particular to a parcel detecting and sorting method and system based on machine vision. According to the method, historical parcel motion trajectory data and real-time video information are fused, a category and weight differentiation trajectory model is constructed, sudden offset is captured in combination with computer vision, the trajectory is accurately predicted, and the attitude offset is obtained. The initial fixed parameter of the actuator is dynamically adjusted to be a first action parameter based on the offset, a compensation coefficient is generated through trajectory tracking error verification, and a second action parameter is obtained; meanwhile, friction coefficient changes are calculated through the real-time track and pressure data, a dynamic response coefficient is determined, and finally a multi-sorting-port cooperation instruction is generated to achieve path dynamic distribution. According to the process, through historical and real-time data fusion, dynamic parameter adjustment, closed-loop feedback and multi-system cooperation, the problems of traditional sorting static presetting, parameter blind adjustment, resource allocation imbalance and the like are solved, and the sorting precision, adaptability and efficiency are improved.

The invention relates to the technical field of man-machine collaboration and information processing, discloses an AI processing-based motion capture recognition pre-judgment system, and aims to solve the problems of difficulty in physiological signal monitoring and performance drifting of a myoelectricity intention recognition model in a high-noise environment, and the AI processing-based motion capture recognition pre-judgment system comprises a processing module, a visual capture module, a myoelectricity acquisition module, a bioacoustics acquisition module and an information prompting module, according to the method, myoelectricity and visual information decoding operation intentions are fused, a self-adaptive filter is guided through the intentions, accurate noise suppression is carried out on biological acoustic signals of a target object, and therefore the physiological state risk of the target object is accurately evaluated, meanwhile, an online self-adaptive calibration mechanism is introduced, the actual action of visual recognition serves as a supervision true value, and the accuracy of the physiological state risk of the target object is improved. And the myoelectricity model is continuously calibrated, so that the accuracy of long-term identification is ensured. According to the method, the decoding intention and the physiological risk level are integrated, the composite cooperation instruction is intelligently generated and fed back to the operator, the situation awareness ability of the operator is enhanced, and the safety and efficiency of man-machine cooperation are improved.

The invention relates to a building fire drill evaluation system and method based on WiFi signal positioning, and belongs to the technical field of rescue drill, and the system comprises a visual capturing module which is used for capturing a drill scene and recognizing rescue workers and rescued persons in the scene; the WiFi signal module is used for capturing WiFi signal changes and analyzing signal reflection characteristics so as to track personnel positions and action changes; the data processing module is used for performing synchronization, fusion and feature extraction on the video data collected by the visual capture module and the WiFi signal data; and the effect evaluation module is used for automatically generating a detailed drill evaluation report, and the report comprises key indexes, behavior analysis and video playback. According to the invention, escape and position conditions of people in a building and rescue conditions of rescue workers can be simulated when a disaster occurs.

The invention discloses an instrument of an intelligent breathing and motion capture system based on six-character table health maintenance, and relates to the technical field of intelligent health maintenance equipment, the instrument comprises a hardware part and a software part; the hardware part comprises a mouth shape visual capture assembly, an action force and angle sensing assembly, an audio acquisition and feedback assembly and a main control and power supply assembly; the software part comprises a data acquisition module, a signal preprocessing module, a parameter analysis and judgment module, a real-time feedback module and a data storage and management module; according to the invention, through integration of the mouth-shaped visual capture assembly, the action force and angle sensing assembly, the audio acquisition and feedback assembly and the master control and power supply assembly, synchronous acquisition and processing of multi-source data are realized; the mouth shape change, the body movement and the audio pronunciation of the user can be accurately captured in real time, and powerful technical support is provided for standardized practice of six-character table health maintenance.

The invention discloses a multi-modal large model training method and an image-text translation method and device, and the method comprises the steps: determining a training sample comprising a training text, a translation text, a training image and inquiry information, training a to-be-trained multi-modal large model based on the training sample, and obtaining a trained multi-modal large model, the multi-modal large model comprises an image encoder, an adapter and a large language model, and the image encoder is used for extracting information with different granularities from pixel representation of an image and converting the information into image vector representation; the adapter is used for converting the dimension represented by the first image vector and the dimension represented by the second image vector into dimensions accepted by a large language model, and the large language model is used for outputting corresponding answers according to the input image vector representation and inquiry information. According to the invention, the fine-grained image encoder is added to the model framework, so that the defect that the multi-modal large model is insufficient in fine-grained image information encoding is overcome, and the visual capture capability of the multi-modal large model is improved.

The utility model discloses a screen vision detection mechanism and screen module detection equipment, the screen vision detection mechanism comprises a vision mounting rack, at least one first camera and at least one second camera, the vision mounting rack is movably arranged, and the first camera and the second camera are both arranged on the vision mounting rack. A view-finding lens of the first camera is vertically downward and faces the front face of the screen, and a view-finding lens of the second camera is obliquely downward and faces the side face of the screen. Through the arrangement of the first camera and the second camera, all-directional visual capture can be performed on the front surface and the side surface of the screen, the comprehensiveness of visual detection is improved, meanwhile, as the visual mounting frame is movably arranged, the first camera and the second camera can perform visual detection on the screens at multiple stations, and the visual detection efficiency is improved. And the detection cost is greatly reduced.

The invention discloses a digital human live broadcast microphone connection and video generation method based on computer vision, and belongs to the technical field of artificial intelligence, and the method comprises the following steps: 1, real-time vision capture and action expression driving: employing a computer vision tool to carry out face key point detection, facial feature points of lips, eyes and the like of a live anchor or a microphone-connected guest are accurately captured; and then a voice conversion model is matched, microphone-connected audio features and lip key point movement are bound, precise synchronization of the sound and the lip is achieved, and for limb actions, limb joint data can be collected through a monocular camera in combination with a posture estimation algorithm or a simple action capture device and mapped to a digital human skeleton model. According to the digital human live broadcast microphone connection and video generation method based on computer vision, through multi-dimensional technical innovation and systematic design, the problems of insufficient sense of reality, interaction lagging, poor scene adaptability and the like existing in current digital human application are effectively solved.

The invention provides a human body motion visual capture and evaluation method and system, and the method comprises the steps: carrying out the video frame decomposition of a human body motion video, and obtaining a plurality of single-frame images; using a first Visual Transform model to detect each single-frame image, and obtaining a 2D human body joint position of each single-frame image; according to the 2D human body joint position of each single-frame image, generating a 2D human body skeleton corresponding to each single-frame image; processing the 2D human body skeleton corresponding to each single-frame image by using a Transform decoder to obtain 3D human body posture data corresponding to each single-frame image; obtaining a 3D human body model corresponding to each single-frame image according to the 3D human body posture data; and comparing the 3D human body model corresponding to each single-frame image with a set standard action to obtain a human body action score. The accuracy of human body action evaluation can be improved.

The application discloses a kind of visual laminating machine, including rack, perspective laminating part, inlet and outlet part, vacuum pump part, vacuum pipeline part, visual capture part, display and electrical control system, perspective laminating part and inlet and outlet part are spaced apart along horizontal direction and are installed on rack and head and tail link;Vacuum pump part is placed in the inside of rack and is located below inlet and outlet part;Vacuum pipeline part is installed in the inside of rack and is located below perspective laminating part, and vacuum pump part is connected with perspective laminating part by vacuum pipeline part and is communicated;Visual capture part is installed on rack and is aligned with perspective laminating part below it;Display is installed on the outside of rack, and visual capture part is electrically connected with display.The application is realized real-time visual monitoring of laminating whole process by the cooperation of visual capture part and perspective upper box, makes process from black box into transparent controllable, quality defect can be found in time and intervened, so as to effectively improve product yield, shorten process verification period, reduce resource waste.

The invention discloses a real-time pair practice training system and method based on a multi-modal large model, and relates to the technical field of large model pair practice. A multi-source sensor array comprising a visual capturing module, an auditory processing module and an action sensing module is deployed, and space-time synchronization of sensor data is realized by using a multi-modal alignment engine; generating a visual space-time vector reflecting a facial dynamic state and a limb time sequence rule; generating a fundamental frequency of the voice signal and an auditory feature vector of formant dynamic change; and generating a pressure sensitive feature vector fusing the hand motion trail and the contact force data. And carrying out weighted fusion on the feature vectors, calculating a comprehensive behavior evaluation index, comparing the comprehensive behavior evaluation index with a preset positive threshold value and a preset negative threshold value, and judging a user behavior state. And calculating a prediction error according to the actual state feedback of the user, and adjusting a positive threshold and a negative threshold in real time by using a dynamic threshold learning updating mechanism to realize adaptive optimization of an evaluation standard.

The invention relates to the technical field of computer vision, and discloses a music playing posture real-time reminding system based on visual capture, which solves the problem that the position of a camera is difficult to flexibly adjust according to different musical instrument types, player heights and training key points at present, and comprises a base, and a bracket is fixedly mounted at the bottom of the base; a first adjusting mechanism is arranged at the top of the base, a second adjusting mechanism is arranged on the first adjusting mechanism, the first adjusting mechanism comprises a screw fixed to the top of the base, a lifting cylinder is arranged on the outer side of the screw, a top plate is fixedly connected to the top end of the lifting cylinder, two guide rods are symmetrically and fixedly connected to the bottom of the top plate, and two guide holes are symmetrically formed in the top end of the screw. The two guide rods are inserted into the two guide holes correspondingly, and a driving motor is fixedly installed at the top of the top plate. According to the invention, the height of the camera can be conveniently adjusted, so that the position of the camera can be flexibly adjusted according to different musical instrument types, heights of players and training key points.

The invention provides a residual vision capture evaluation method and device, a medium and a program product, and relates to the field of ophthalmology detection. According to the method, a fixation point is set in a virtual scene and used for focusing the visual field of a user; monitoring an interaction operation of a user on the virtual scene, when the interaction operation is detected, controlling a first line segment to appear in a first direction of the virtual scene, and moving the first line segment from the first direction along a direction opposite to the first direction at a first speed; and when the first line segment breaks or recovers when the first line segment moves along the moving path at the current position, monitoring and responding to interactive operation for respective current positions during breaking or recovering, and recording current position data of the first line segment. The method and the device are easy and convenient to operate and high in portability, the visual field detection device and method do not need assistance of professionals, a detector can conduct visual field detection autonomously and conveniently, and the requirement for visual field measurement can be completely met on the community screening level.

The application provides a humanoid robot facial expression mapping and calibration method, the facial expression mapping method constructs a self-supervised discrete expression data set, trains a multilayer perceptron model based on the data set, realizes preliminary mapping from expression parameters to rudder control signals, extracts expression parameters in a real person expression data set, generates robot rudder control signals through the model, constructs a self-supervised time sequence expression data set based on a redirection method, trains a long short-term memory network model, and realizes time sequence mapping of continuous expressions. The facial expression calibration method uses a visual capture device to obtain robot facial expression parameters in real time, gradually optimizes the rudder control signal, and makes the expression parameter response approach the target value. Through data-driven modeling and visual feedback optimization, the application realizes high-fidelity mapping from semantic expression parameters to multi-rudder collaborative control and systematic calibration, effectively improving the naturalness, accuracy and long-term stability of the robot facial expression.

The invention belongs to the technical field of underwater cable vibration measurement, and discloses an underwater cable three-dimensional vortex-induced vibration experiment device and method based on visual measurement, and the device comprises a wave flow water tank, a wave and current generator, a cable model, a target spot, and a data collection device. The current generator is arranged at one end in the wave flow water tank, the cable model is arranged in the outlet direction of the current generator, the target spot is arranged on the cable model, and the data acquisition device is arranged on the side face of the cable model and used for acquiring vibration data of the cable model. According to the method, the position of the cable model is calibrated by setting the target spot, the coordinate change of the target spot is captured by the visual capture device, visual measurement data can be extracted by using a computer program, and the method is not in contact with the cable model and does not influence a flow field around the cable model. Meanwhile, the visual catcher is immersed in water, light refraction is avoided, measurement errors are reduced, and the result is more accurate.

The application provides a residual visual capture evaluation method, device, medium and program product, and relates to the field of ophthalmic detection. The method sets a fixation point in a virtual scene to focus the user's visual field; monitors the user's interactive operation on the virtual scene, controls a first line segment to appear from a first direction of the virtual scene when the interactive operation is detected, and the first line segment moves from the first direction along a direction opposite to the first direction at a first speed; when the first line segment breaks or recovers during movement along the movement path at the current position, the interactive operation for the respective current position when the break or recovery occurs is monitored and responded to, and the current position data of the first line segment is recorded. The method and device of the application are simple to operate, strong in portability, and do not require a visual field detection device and method assisted by professionals, so that the detector can independently and conveniently perform visual field detection, and the demand for visual field detection at the community screening level can be fully met.

The invention relates to the technical field of visual capture, in particular to an intelligent autonomous navigation and positioning system with a body based on machine vision. The system comprises an obstacle avoidance route state analysis module and a state evaluation module. According to the invention, an obstacle avoidance route action sequence is formulated through an obstacle avoidance route state analysis module, a corresponding obstacle avoidance route is allocated according to the movement sequence of the transportation equipment, information collection is carried out by utilizing the transportation equipment to actually pass through the obstacle avoidance route, and the obstacle avoidance state corresponding to the transportation equipment acting on each obstacle avoidance route is collected; and the state component evaluation module is matched with the obstacle avoidance state of the obstacle avoidance route to evaluate the state component corresponding to the obstacle avoidance route, and the obstacle avoidance route is selected through the state component so as to adapt to the type of transported goods of actual transportation equipment, match the obstacle avoidance route meeting the actual demand for different types of transported goods, and improve the adaptation effect of the obstacle avoidance route.

This application relates to the field of navigational lighting, and more particularly to a method, system, and storage medium for detecting contamination at the light outlet of a navigational lighting fixture. The method comprises real-time visual capture of the light outlet surface of the navigational lighting fixture to obtain visual data; determining contamination data by comparing the visual data with preset visual threshold data; and comparing the contamination data with preset contamination threshold data. If the contamination data exceeds the threshold data, a corresponding alarm or indication is generated to a worker. This application improves maintenance efficiency.

The present application relates to the technical field of video monitoring, and discloses a camera cooperative monitoring method and system based on multi-modal perception, which comprises detecting abnormal event signals in the environment through a non-vision sensor node, triggering camera activation and visual capture of a target area; the camera performs relay tracking based on target feature matching to generate a continuous motion trajectory; multi-view cooperative shielding is performed on privacy-sensitive information in the tracking target to generate desensitized monitoring data and upload the same; camera energy consumption is optimized during a non-event triggering period, and a behavior recognition model is updated based on the desensitized data. The method solves the contradiction between privacy protection and monitoring efficiency, realizes cross-region model evolution and energy consumption reduction, and systematically avoids the problems of traditional monitoring response lag, data redundancy, privacy leakage and high energy consumption.

The application relates to the computer technical field and discloses a steel structure intelligent visual defect detection method and system, aiming to solve the technical problem that it is difficult to find hidden defects such as closed micro-cracks by simply relying on visual detection means in the prior art. The method comprises the following steps: performing sweep excitation of a specific frequency range on a steel structure to make the defect area produce a micro-vibration response; using a high-frame-rate industrial camera combined with Lagrange video amplification technology to capture the sub-pixel level vibration amplitude; and outputting a defect detection report based on a support vector machine classifier. The system comprises an acoustic vibration excitation module, a high-frame-rate visual capture module, a cross-modal fusion analysis module and a defect detection report output module. By combining acoustic vibration excitation and visual capture, the application can effectively activate and detect hidden defects, realize non-contact, rapid, high-sensitivity intelligent defect identification, and significantly improve the detection accuracy and reliability.

The application provides a human action visual capture and evaluation method and system, wherein the method comprises: performing video frame decomposition on a human action video to obtain a plurality of single-frame images; using a first Visual Transformer model to detect each single-frame image to obtain the 2D human joint position of each single-frame image; generating the 2D human skeleton corresponding to each single-frame image according to the 2D human joint position of each single-frame image; using a Transformer decoder to process the 2D human skeleton corresponding to each single-frame image to obtain the 3D human posture data corresponding to each single-frame image; obtaining the 3D human model corresponding to each single-frame image according to the 3D human posture data; and comparing the 3D human model corresponding to each single-frame image with a set standard action to obtain a human action score. The application can improve the accuracy of human action evaluation.

The invention discloses a sea rescue system which comprises a flying disc type life buoy and a life buoy catching device, the flying disc type life buoy comprises an annular floating body and a water-soluble film system, and the water-soluble film system wraps the whole annular floating body; the water-soluble film system is provided with an inflation inlet in the range of the inner ring of the annular floating body, and the longitudinal section of the annular floating body is composed of two wing-like sections in mirror symmetry; and the life buoy capturing device is used for pulling back the life buoy. According to the sea rescue system provided by the invention, the life buoy adopts an asymmetric section structure and a water-soluble film system, so that efficient remote accurate throwing is realized, the aircraft system quickly positions the position of the life buoy through the visual capture system, the mooring rope is brought to the position of the life buoy, and the winch quickly tows the life buoy and personnel to a ship; rapid and long-distance rescue of the ship is achieved, rescue cost is reduced, and rescue efficiency is improved.

The invention relates to the technical field of wire defect detection, and discloses a flat electromagnetic wire online defect detection system which comprises an equipment outer frame, a detection table movably mounted on the equipment outer frame and visual capture equipment located on the equipment outer frame. The visual capture equipment comprises an installation tool located on the front face of the equipment outer frame and a CCD camera fixedly installed at the bottom of the installation tool, a light source structure used for supplementing light to the CCD camera is arranged on the installation tool, and a groove is formed in the outer side of the installation tool. According to the flat electromagnetic wire online defect detection system, by arranging the point light source for supplementing light to the CCD camera, annular light supplementing lamps installed on the outer side of the installation tool can be enhanced, the shooting light supplementing effect of the CCD camera is better, meanwhile, the multiple light supplementing lamps are rotationally adjusted by arranging the adjusting plate, and the detection accuracy is improved. The light supplementing lamp is convenient to adjust and use.

The invention discloses a multi-modal emotion recognition method based on image and character early fusion. The method comprises the following steps: preprocessing a multi-modal emotion data set to obtain a data sequence of the multi-modal emotion data set; establishing an interactive fusion model based on image and character multi-modal emotion data by using a data sequence of the multi-modal emotion data set; performing multi-sample loss training on an interactive fusion model based on image and character multi-modal emotion data by using a data sequence of the multi-modal emotion data set; and inputting test data by using the interactive fusion model based on the image and character multi-modal emotion data to obtain a multi-modal emotion recognition result, and analyzing model performance. According to the method, the ability of visual capture and context information understanding is improved, the model is allowed to adapt to knowledge in a specific field, generalization is improved through early fusion, and the method is superior to an existing method in the aspects of sentiment analysis and the like.

The utility model discloses a manipulator clamping jaw for inferior-quality product grabbing and capable of early warning, which comprises a fixed shaft rod, a clamping assembly and a visual capturing assembly, the clamping assembly is arranged at one end of the fixed shaft rod, the clamping assembly comprises a driving part, a moving seat, a clamping jaw and a fastening end, the driving part is arranged at one end of the fixed shaft rod, and the moving seat is arranged at the other end of the fixed shaft rod. The movable seats are arranged at the output end of the driving part respectively, the clamping jaws are arranged on the movable seats correspondingly, the fastening ends are detachably arranged on the clamping jaws correspondingly, and the visual capturing assembly is arranged on the fixed shaft rod. Therefore, when the output flange is grabbed, the position of the output flange is judged through visual capture, so that whether the output flange can be grabbed by a manipulator or through a clamping jaw or not is judged, if the output flange cannot be grabbed, an operator is warned in time to carry out manual processing, and therefore the operation process of the whole production line is optimized, and efficiency is improved.

This invention relates to a system and method for constructing a multimodal human behavior perception dataset based on CSI (Computer-Sensitive Interface). The system includes a wireless sensing subsystem for acquiring first CSI data packets of the human body, a visual capture subsystem for acquiring first RGB video, first depth video, second RGB video, and second depth video, a Linux-based central control unit, and a multimodal post-processing unit for cleaning and enhancing 3D skeletal joint data, denoising and extracting features from CSI data, and generating hierarchical natural language text descriptions. This invention constructs a large-scale, multidimensional, high-fidelity human behavior perception dataset acquisition system by deeply integrating high-frequency, fine-grained wireless signals with visual and semantic ground truth, ensuring the logical coupling of high-dimensional CSI signals and visual data.

The invention relates to the technical field of image processing, in particular to a robot arm vision capturing method and system, and the method comprises the following steps: continuously obtaining an infrared image, a color image and a distance image of a captured object, and obtaining a robot arm motion image of a robot arm; based on the infrared image, the color image and the distance image, the robot arm uses an improved single-stage detection network to recognize the grabbed object; after the grabbed object is recognized, a plurality of feature points are marked on the grabbed object and the manipulator, and the color image and the manipulator motion image are used for tracking and positioning the feature points; and the positions of the grabbed object and the manipulator are determined by tracking and positioning the feature points. According to the method, the grabbed object is identified based on the infrared image, the color image and the distance image of the grabbed object and the robot arm motion image of the robot arm, so that the accuracy and the reliability of visual capture are improved.

The invention relates to a medical image segmentation method based on edge enhancement and multi-feature fusion. The method comprises the following steps: extracting multi-scale local features of an image through an encoder; in an edge enhancement module of a bottleneck layer, boundary information under a global view angle is captured in combination with differential convolution and visual MLP, so that accurate transmission of edge details in low-layer features is ensured; meanwhile, through a multi-feature fusion module embedded in jump connection, high-level semantics and low-level details are integrated, mask information is introduced to assist feature alignment, and the local structure and the overall context of the image are considered; then, the decoder gradually recovers the spatial resolution and strengthens the boundary feature response. The method can generate an accurate segmentation mask, effectively reduces the interference of a complex background on a segmentation result, improves the segmentation precision of an algorithm on a focus region, can effectively cope with the challenges of blurred tissue edges, focus scale change and the like in a medical image, remarkably inhibits background noise interference, and improves the segmentation accuracy. Therefore, the segmentation precision and the boundary fitting degree are better.

The application discloses a kind of injection drug identification methods based on machine vision, specifically comprising: the drug to be identified is placed in visual identification area;Rotating the drug to be identified and / or camera unit to switch the orientation of the feature identification point of the drug to be identified relative to the camera unit, so that the feature identification point of the drug to be identified falls into the visual capture range of the camera unit;The feature identification point image of the drug to be identified is captured by the camera unit;The drug information of the drug to be identified is identified according to the feature identification point image captured by the camera unit.The application also discloses an injection drug identification system and an intelligent dispensing equipment.The injection drug identification method, system and intelligent dispensing equipment provided by the application ensure that the feature identification point of the drug to be identified can be successfully captured by the camera unit for drug information identification by rotating the drug to be identified and / or the camera unit, and this identification method only needs to set a single camera unit, which can save cost and space.

This application relates to an electroencephalogram monitoring system for neurology patients. The system includes a signal acquisition module, a visual capture module, and a signal processing module, where: the signal acquisition module is used to collect spontaneous bioelectrical potentials of the user's head and output the original signal to the signal processing module; the visual capture module is used to collect the user's eye movement information, and the eye movement information includes the user's blink movement information, and the visual capture module outputs the eye movement information to the signal processing module; the signal processing module is used to receive the eye movement information and the original signal, segment the original signal according to the blink movement information in the eye movement information, and perform blink artifact removal processing on the partial original signal segments containing the blink movement information after segmentation.