513 results about "View camera" patented technology

The invention relates to the technical field of image processing, and discloses an industrial part alignment method and system based on visual analysis and a storage medium. The method comprises the steps that a three-view camera collects an industrial part image, and preprocessing is carried out through gradient magnitude local contrast enhancement to obtain an enhanced image; performing hierarchical feature extraction to identify edge contours and key control points to form a multi-dimensional feature set; and establishing a dynamic reference coordinate system based on the feature set to obtain a part space attitude matrix. And the attitude deviation is compensated through Z-axis offset and rotation coupling error analysis. Posture adjustment is decomposed into a plurality of sub-stages, an alignment track is optimized by adopting a variable speed planning strategy, and accurate alignment of the parts is achieved. The problems that multi-view visual information fusion is insufficient, a special recognition algorithm for geometrical characteristics of the industrial parts is lacked, and Z-axis offset and rotation coupling error compensation is inaccurate in the posture adjustment process are solved, and the precision and stability of alignment of the industrial parts are improved.

The invention provides an intelligent heavy truck end-to-end driving method based on focus attention and probabilistic game, and belongs to the technical field of intelligent driving. The method comprises the following steps: firstly, acquiring multi-source data such as a laser radar, a look-around camera, an IMU (Inertial Measurement Unit) and a load signal, and generating an anti-jitter space-time BEV feature through cross-dimensional position coding based on motion compensation, near-field non-uniform sampling, focus attention and time sequence fusion; then, vectorization decoding is carried out on the features to obtain obstacle vehicle movement and map element vector information. And finally, on the basis of the expert track prior space, probability game interaction is carried out through multi-round iterative prediction and planning, a self-vehicle track conforming to heavy truck variable load dynamics constraints is generated, and the self-vehicle track is output after safety verification. According to the method, the problems of heavy truck high-position sensor vibration distortion, hinge structure blind area and trajectory planning under the large-inertia variable-load working condition are effectively solved, and the driving safety and robustness are remarkably improved.

The invention belongs to the technical field of laser radar-camera panoramic segmentation, and particularly relates to a multi-mode panoramic segmentation method for multi-view space-time alignment and implicit feature interaction, and the method is executed by a multi-mode panoramic segmentation network, and comprises the steps: S1, obtaining laser radar point cloud data and multi-view camera image data in the same scene; s2, performing double-branch feature coding on the laser radar point cloud data; performing multi-scale image feature extraction on the camera image data; s3, generating false point cloud features with geometric perception capability; s4, generating semantic pixel features; s5, performing implicit fusion on the pseudo point cloud features generated in the S3 and the semantic pixel features generated in the S4 to obtain cross-modal fusion features; and S6, based on the cross-modal fusion features obtained in the S5, generating a unified panoramic segmentation result containing semantic tags and instance IDs. The method can effectively improve the robustness and precision of multi-mode panoramic segmentation in a complex urban environment.

The invention discloses a tobacco logistics carbon emission real-time monitoring method and system based on AI visual identification. According to the method, vehicle operation videos and equipment working state images in a logistics park are collected in real time through a multi-view camera, sensor data are synchronously obtained, and multi-source data fusion is achieved through a time alignment and normalization method. A space-time fusion network model is constructed, the idle speed state of a fuel vehicle is detected by using improved YOLOv7, equipment energy consumption time sequence characteristics are analyzed in combination with Transform, and a space correlation model between equipment is established through a graph convolutional network. And inputting the extracted features and power data into a TCN-LSTM-Attention model subjected to Bayesian optimization, calculating real-time carbon emission, and dynamically adjusting model parameters based on a result. And finally, the edge-cloud collaboration platform generates hierarchical alarms and feeds back regulation and control instructions to form a closed-loop control loop. According to the invention, accurate monitoring and intelligent optimization of carbon emission in the whole process of tobacco logistics are realized, and an innovative solution is provided for green logistics.

The application relates to the technical field of blind area monitoring, and relates to a low-speed blind area monitoring method and system based on multi-view camera shooting, which comprises the following steps: registering multiple cameras in a vehicle to obtain a target vehicle and a vehicle coordinate system; confirming multiple early warning coordinates based on the vehicle coordinate system, a vehicle bird's eye view and the vehicle; confirming a candidate window based on the vehicle bird's eye view and the early warning coordinates; performing target detection on the candidate window by using a pre-constructed target detection model to obtain a detection result; if the detection result is a first result, taking a pre-constructed safety instruction as a target instruction; otherwise, confirming relative positions and relative speeds based on target bounding boxes in a second result and multiple radars in the target vehicle; calculating an expected collision time according to the relative positions and the relative speeds; confirming the target instruction based on the expected collision time; and summarizing the target instruction to obtain multiple target instructions, thereby completing low-speed blind area monitoring. The application can improve the precision of low-speed blind area monitoring.

The invention relates to the technical field of intelligent traffic and artificial intelligence, and discloses an end-to-end automatic driving method and system based on multi-modal attention fusion, and the method comprises the steps: collecting multi-modal data; generating an initial semantic text and calibrating the initial semantic text to form a calibrated semantic text; performing feature processing on the environment image and the point cloud data acquired by the multi-view camera, performing cross-modal alignment on the calibrated semantic text and the first BEV feature, and performing spatial-temporal feature modeling; generating candidate trajectories, quantifying the collision risk of each candidate trajectory and the dynamic obstacle, screening out low-risk trajectories, and guiding the low-risk trajectory optimization through a rule mask; mapping the optimized track into a control instruction; according to the method, a CLIP cross-modal alignment mechanism is utilized, text features and BEV geometric features are deeply fused, and the recognition accuracy of the key region is improved.

The invention belongs to the technical field of industrial robot motion control, and discloses an industrial robot precise path generation system based on three-dimensional visual reconstruction, comprising an environment sensing device which is composed of a multi-baseline structured light projector and a multi-view camera array and obtains point cloud data through light interference suppression and spectrum correction; the three-dimensional reconstruction processing unit generates a three-dimensional model through rigid transformation and deep learning correction; the geometric feature analysis unit performs curvature analysis and boundary recognition to form region description; the trajectory planning device adopts a weighted B spline to generate a trajectory under curvature constraint and energy optimization; the dynamic correction unit is switched between vision feedback and force feedback according to an error threshold value; the control conversion processor combines inverse kinematics and dynamics compensation and predicts deviation to output a control instruction; and the execution driving mechanism utilizes a servo motor and closed loop feedback to execute actions, so that high-precision generation and stable execution of a track under a complex working condition are realized.

The invention discloses an intelligent multi-view camera tracheal intubation image processing method, device and system, and relates to the technical field of medical instruments, the method comprises the following steps: in a tracheal intubation process, a multi-view image acquisition unit collects a plurality of images and combines the images to form an image pair; and carrying out stereo matching and depth calculation on the image pairs to generate a depth map. And performing three-dimensional reconstruction by combining the depth map and the original image to generate a three-dimensional model. And finally, identifying the target in the three-dimensional model and outputting the depth of the target as an image processing result. The technical problems that an existing multi-view camera is inaccurate in depth perception and insufficient in real-time performance during tracheal intubation image processing are solved, and the technical effect of improving the depth perception precision and the real-time processing capacity of the multi-view camera during tracheal intubation image processing is achieved.

The invention relates to the technical field of image shielding processing, in particular to a target graph shielding processing method and system based on variable-focus wide-view-angle cameras, and the method comprises the steps: deploying a plurality of variable-focus cameras based on seat distribution, constructing a monitoring network, and collecting an initial image; identifying desktop features according to the images, establishing a unified coordinate system and mapping each desktop; when shooting is needed, querying candidate cameras according to the coordinate system, and determining a main camera according to an imaging area and a distance score; and performing quality evaluation on the main shot image, outputting the main shot image if the main shot image reaches the standard, otherwise, triggering a dynamic re-shooting mechanism. According to the invention, through cooperation of multiple cameras and intelligent visual angle selection, the problem that a single camera is easy to block is effectively solved, and automatic and high-quality image acquisition and processing are realized.

The invention provides an automatic driving target detection method and electronic equipment, and belongs to the technical field of computer vision and automatic driving, and the method comprises the steps: obtaining multi-view image data through a vehicle-mounted multi-view camera, carrying out the preprocessing of the image data, and inputting the image data into a backbone network to extract multi-scale image features; performing channel attention weighting and space attention weighting on the multi-scale image features, and inputting the weighted features into a bidirectional feature pyramid network for bidirectional fusion to generate enhanced multi-scale fusion features; generating auxiliary target category prediction and bounding box prediction by using an auxiliary prediction head based on the multi-scale fusion features, and calculating auxiliary supervision loss according to the target category prediction and bounding box prediction so as to apply direct supervision to the backbone network; and integrating the obtained auxiliary prediction bounding box into a position query of a decoder to form a mixed query, inputting the mixed query and BEV features into the decoder for optimization, and outputting a final target detection result.

Systems and methods for navigating a host vehicle are disclosed. In one implementation, a system includes a processor configured to receive at least one image captured by a front view camera of the host vehicle; analyze the at least one image to detect a representation of a traffic sign in an environment of the host vehicle; determine a relevancy of the traffic sign to the host vehicle, wherein the relevancy of the traffic sign to the host vehicle is determined based on a trajectory of the host vehicle relative to a location of the traffic sign; and cause the host vehicle to implement at least one navigational action based on the relevancy of the traffic sign to the host vehicle.

The invention provides a three-dimensional wave field measurement analysis method and system based on a multi-view camera, and relates to the technical field of data measurement, and the method comprises the steps: collecting a wave field multi-angle image through the multi-view camera, building a space projection relation, obtaining a feature point set, and reconstructing an initial three-dimensional point cloud; extracting time-varying features by adopting a dynamic feature learning network; constructing a deformation constraint rule for correction according to a fluid mechanics equation; and finally calculating the three-dimensional characteristic parameters of the wave field. According to the invention, high-precision three-dimensional reconstruction of the wave field can be realized, the measurement accuracy is improved, and reliable technical support is provided for hydraulic engineering and marine environment monitoring.

A method of operating a surround view camera system for a vehicle includes generating first image data at a first time using an imaging device, the first image data corresponding to a first image of a surroundings of the vehicle, and receiving first vehicle data generated by at least one sensor of the vehicle with a processor, the first vehicle data generated at the first time. The method further includes generating second image data at a second time after the first time using the imaging device, the second image data corresponding to a second image of the surroundings of the vehicle, and receiving second vehicle data generated by the at least one sensor with the processor, the second vehicle data generated from the first time to the second time. The method also includes processing the first vehicle data and the second vehicle data using the processor to determine change data.

The invention relates to a multi-view video real-time splicing method and device, belongs to the technical field of multi-view video splicing, and solves the problems of insufficient view fields in the vertical direction and the horizontal direction during three-view or four-view video splicing, insufficient real-time performance during multi-view video splicing, low splicing quality and high hardware cost. The method comprises the following steps: acquiring frame images respectively shot by a multi-view camera in real time; wherein the multi-view cameras are uniformly distributed on a 180-degree arc on the same horizontal plane, and each optical axis is perpendicular to the horizontal plane; based on the image remapping parameter of each camera, mapping pixel points of each frame of image to a target coordinate system, and generating a seam mask between adjacent images; wherein the adjacent images are frame images shot by two adjacent cameras; and inputting the frames of images unified to the target coordinate system and the seam masks into a fusion device to obtain a spliced panoramic image.

The invention relates to a camera parameter adjusting method and device, a vehicle, a medium and a program product, and is applied to the vehicle, and the adjusting method comprises the steps: determining a target adjusting and optimizing mode based on the information of a road where the vehicle is located; determining a detection result based on the current driving information and the current lane line information of the vehicle; wherein the detection result represents whether parameter adjustment is triggered or not; under the condition that the detection result is a first detection result, according to a target adjustment and optimization mode, determining current adjustment information of the first surround view camera based on a target image which is acquired by the first surround view camera and contains a dotted line lane and previous adjustment information of the first surround view camera; and according to the current adjustment information of the first around-view camera, adjusting the external parameters of the first around-view camera. According to the invention, the adaptability and accuracy of parameter adjustment of the around-view camera of the vehicle can be improved.

The invention provides a vehicle sentry mode environment sensing method and system and electronic equipment, and the method comprises the steps: carrying out the intrusion judgment of a moving target around a vehicle based on a vehicle sensing system, and controlling the vehicle to enter an all-round sensing working state if there is a moving target crossing a set intrusion line; in the all-round perception working state, grounding information of all the moving targets is obtained based on an all-round camera, and corresponding position coordinates are calculated and obtained according to the grounding information; and performing sentry mode early warning based on the position coordinates of the moving targets. According to the invention, after the vehicle sensing system detects intrusion of the moving target, the vehicle is controlled to enter the surround-view sensing working state, target detection and distance measurement are performed on the moving target through the surround-view camera so as to obtain the behavior track of the moving target, false alarm and false video recording in a sentry mode are reduced, and the safety of a sentry is improved. And meanwhile, the computing power of all-round parking perception of the shared vehicle reduces the computing power and data cost of sentry mode environment perception.

The invention relates to a deep feature fusion-based fabricated component surface defect multi-task detection method. The method comprises the following steps of: constructing a multi-view camera array and an illumination feedback regulation and control module at a data acquisition end; on the algorithm level, feature cross-layer propagation and information compensation are realized through an improved lightweight convolutional network and a multi-scale residual diffusion module; a dynamic feature fusion module is introduced, and a self-adaptive fusion weight is generated based on channel statistical features, so that feature sharing and differential expression are realized among different tasks; meanwhile, a defect perception attention mechanism and cross-task consistency constraint are adopted, and the problem that semantic space distribution is inconsistent in the multi-task detection process is solved; in the detection post-processing stage, a three-dimensional quantitative evaluation system based on the geometric dimension, the texture roughness and the depth volume is constructed, and unified grade evaluation of the surface defects of the component is achieved through the multi-feature fusion quality index. The problems of low detection efficiency, unstable precision, difficulty in collaborative recognition of multiple types of defects and the like in existing component delivery detection are solved.

An interactive gaming system comprising a multiplicity of game access points used by guests to engage with an on-going destination game. Game access points comprise traditional destination activities as well as additional game parts. Gaming guests carry one or more electronic means and are tracked throughout the destination locations and activities. At least one electronic means comprises a visual display for receiving requested directions or a system summons to a game access point. The electronic means comprises a direct view lens system such that the guest looks directly through the electronic means to a destination map or destination scene, where directions are overlaid onto the direct view and registered to the map or scene. The mobile device further comprises a front view camera for capturing images of the guest's direct view of a destination game part, where the image is used as actionable game information.

The invention discloses a transformer substation video fusion monitoring method and system based on 3DGS, and relates to the technical field of three-dimensional modeling. The method comprises the following steps: acquiring a laser radar point cloud and a multi-view camera image of a transformer substation, and constructing a three-dimensional Gaussian splashing model; establishing a mapping relation library based on P source view angle and Q target view angle transformer substation 3DGS samples; based on the library, using a picture splicing boundary confidence degree adjusting pointer and a space coordinate mapping deviation correction pointer to carry out fusion training; and when a user clicks a target device in the model, calling an improved line-of-sight cone analysis algorithm to dynamically screen an optimal camera, receiving a video stream of the optimal camera, and determining a device state recognition result. The technical problem of poor accuracy of image splicing and equipment positioning caused by multi-view and view range limitation in substation equipment monitoring is solved, and the technical effect of improving the monitoring precision and efficiency of substation equipment by dynamically screening the optimal camera and equipment state recognition of space coordinates is achieved.

The invention relates to the technical field of rubber detection, in particular to a rubber part quality detection device and a detection method thereof. The device comprises a detection index plate, a support and two lifting devices are sequentially arranged on the periphery of the detection index plate, a laser diameter measuring sensor is arranged on the support, and the laser diameter measuring sensor is arranged above a vibration disc feeder in a suspended mode and used for measuring the diameter of a passing rubber part. A top view camera and a side view camera are connected to the lifting device, an annular lamp is fixedly arranged on the periphery of a lens of the top view camera, a supporting rod is fixedly connected to the lifting device corresponding to the side view camera, and a parallel backlight lamp is arranged at the end of the supporting rod; light shields are arranged on the peripheries of the annular lamp and the parallel backlight lamp, baffles are movably arranged in the light shields, and driving assemblies are connected between the baffles and the light shields; the driving assembly adjusts the rotation angle of the laser diameter measuring sensor according to detection data of the laser diameter measuring sensor, drives the baffle to move in the light shield, and adjusts the local shielding range of the baffle to the light emitting area of the corresponding light source.

The invention relates to an AprilTag-based panoramic camera 3D calibration method and system and a medium, and the method comprises the steps: M1, obtaining the original images of a front vehicle-mounted fisheye camera, a rear vehicle-mounted fisheye camera, a left vehicle-mounted fisheye camera and a right vehicle-mounted fisheye camera, carrying out the fisheye distortion correction of each image, and obtaining the data information of a distortionless image; and M2, based on the data information of the distortionless image, detecting an AprilTag mark in each distortionless image, extracting a detected angular point coordinate as an image point set, querying a three-dimensional world coordinate corresponding to a predefined calibration plate physical size, constructing a matched three-dimensional point set, and obtaining the data information of the matched three-dimensional point set of the image. According to the method, flexible adjustment of the height and view parameters of the aerial view camera is achieved, repeated calibration is avoided, the parameters of the virtual camera are dynamically corrected by fusing the data of the vehicle attitude sensor, and view angle compensation under the vehicle body inclination state is achieved.

The invention relates to the technical field of automatic driving, in particular to a multi-mode automatic driving training method based on a DeepSeek training framework. Comprising the following steps: reading multi-view camera images and text instructions of a DriveLM-nuScenes data set, and splicing the images according to a look-around layout to form panoramic representation; performing zooming, normalization and standardization processing on the panoramic image to obtain an image tensor; performing marking processing on the text instruction, inserting an image placeholder and a dialogue role mark, and structuring text input representation; dimensionality alignment, position code addition and cross-modal attention fusion of vision and text marking sequences are realized through a multi-modal alignment module, and multi-modal embedding representation is generated; and inputting the embedded representation into a DeepSeek language model to generate a decision text through autoregression, and taking the cross entropy loss with a mask as an optimization target. According to the method, the problems of insufficient multi-view fusion, weak modal alignment and the like in the prior art are solved, the cognitive reliability and the decision interpretability in a complex scene are improved, and vehicle-mounted edge deployment is adapted.

The invention discloses a laser-point-cloud-guided tunnel multi-view area-array camera distribution map generation method and system, and the method comprises the steps: carrying out the calculation of a spatial mapping relation between a synchronously collected multi-view image and corresponding laser radar point cloud data based on the calibration parameters of a camera and a radar; geometric correction is carried out on the original image, and brightness adaptive equalization preprocessing is carried out on the corrected image; carrying out image feature point extraction and matching on the ROI of the image based on a deep learning model; recording an image matching relationship in a metadata form and creating an index; and based on the constructed image matching relationship metadata and indexes, realizing rapid generation of a panoramic distribution map. According to the method, the generated distribution image splicing error caused by factors such as uneven image brightness and perspective distortion can be relieved; large-scale image data matching and fusion based on adjacent image coincidence ROI constraint significantly improve efficiency; and based on a lightweight deep learning model, the reliability of extracting and matching the feature points of the low-illumination and weak-texture image of the tunnel scene is improved.

The invention discloses a multi-view camera splicing monitoring system based on a sky target, and relates to the technical field of visual intelligence. The multi-view camera splicing monitoring system based on the sky target comprises the following steps: a multi-view data acquisition module used for acquiring multiple types of data sets and carrying out time alignment and dimensionless normalization; the target detection and luminosity reference construction module is used for carrying out sky target detection and standard color block analysis and constructing a multi-view luminosity reference library; the deviation estimation and mapping construction module is used for constructing a cross-camera global luminosity mapping model; the adaptive control and real-time correction module is used for regional luminosity correction and illumination change detection; and the panoramic stitching and evaluation feedback module is used for constructing a joint area luminosity consistency evaluation index. According to the method, the multi-view splicing brightness color consistency and the target detection robustness under complex illumination are effectively improved, and the problems of seam brightness gradient and difficult color cast elimination caused by independent automatic exposure and white balance are solved.

Rear-view image(s) of a region of a surrounding environment that is behind a vehicle, is / are captured, by utilising rear-view camera(s). An image to be displayed via the heads-up display, is generated, wherein when generating the image, processor(s) is / are configured to generate an image segment of the image by utilising the rear-view image(s) of said region of the surrounding environment. The image is displayed via the heads-up display for producing a synthetic light field augmenting a real-world light field incoming via a windshield of the vehicle.

The application relates to the technical field of intelligent auxiliary reminding, in particular to an auxiliary reminding method based on a smart wearable device, which comprises the following steps: collecting behavior data of the elderly through a multi-view camera, an infrared sensor and a pressure sensing device, constructing a four-dimensional characteristic behavior sequence graph, and using a deep learning model with a multi-modal fusion and attention enhancement mechanism to analyze behavior semantics and generate graded reminding instructions. The application can accurately monitor the daily behavior of the elderly, dynamically adjust the reminding strategy, trigger an emergency alarm in an abnormal situation and record intervention logs, and effectively improve the life safety of the elderly and the monitoring efficiency.

1. The name of the design product: automobile traveling data recorder (YKC-AEB101). 2. The use of the design product: the design product is used for commercial vehicle auxiliary driving control, built-in double hardware system, and product functions integrate millimeter wave radar, ADAS, front view camera, DMS, BSD, AEBS, alarm display, display screen, brake actuator, etc. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective view 1.

The invention relates to the field of camera calibration, in particular to a calibration method, device and equipment for a yaw angle of a vehicle foresight camera and a medium. The method comprises the following steps: acquiring a driving direction information sequence of a vehicle in a straight driving state; according to the driving direction information sequence, obtaining a reference driving direction of the vehicle driving along a straight line; determining the actual driving direction of the vehicle at the target moment based on the driving direction information at the target moment in the driving direction information sequence; determining a compensation yaw angle according to the deviation between the actual driving direction and the reference driving direction; and on the basis of the compensation yaw angle, calibrating an estimated value of the external parameter yaw angle of the foresight camera at the target moment, and obtaining a calibration value of the external parameter yaw angle of the foresight camera.

The invention relates to the technical field of image processing, in particular to a bogie part image recognition and positioning method and system based on deep learning, and the method comprises the steps: constructing a bogie part and bolt image data set, and carrying out the preprocessing and marking; a recognition positioning model (introducing an ECA attention mechanism, optimizing a BiFPN feature fusion structure and designing a special bolt detection head) is constructed based on an improved YOLOv8 algorithm, after the model is trained and optimized, images are collected through a multi-view camera and input into the model to obtain two-dimensional coordinates, the two-dimensional coordinates are converted into three-dimensional space coordinates in combination with camera calibration and binocular vision, and the three-dimensional space coordinates are obtained. And after post-processing optimization, transmitting to a robot control system to guide full-automatic decomposition. According to the method, the recognition accuracy and positioning precision of the bogie parts and the bolts in the complex environment are improved, the bolt positioning error is within + / -0.5 mm, the recognition accuracy is larger than or equal to 98%, and support is provided for intelligent bogie maintenance.