252 results about "Pedestrian detection" patented technology

The invention discloses a multi-pedestrian target detection method based on a YOLOCDG network in a complex scene, and aims to solve the problems of small target missing detection, model redundancy and inaccurate positioning in multi-pedestrian detection in a scene with dense people flow and serious shielding. The method comprises the following steps: firstly, based on YOLOv8, deleting a redundant convolution fusion layer of a backbone network, adjusting the size of a detection head, and simplifying a framework to reduce small target feature dilution; secondly, designing a context attention module CoAM and embedding the CoAM into a backbone network, and enhancing feature distinction degree in a shielding scene by capturing cross-target context association of dense pedestrians; thirdly, an improved C2fG module is proposed to replace a neck C2f module, the parameter quantity is reduced, and the detection performance and the edge deployment efficiency in a complex scene are balanced; then, an SPPFDSC multi-scale fusion layer containing depth separable convolution is designed, an original SPPF layer is replaced, and fine-grained feature perception of small-size pedestrians in the distance is enhanced; and finally, optimizing a bounding box loss function by adopting WIOU v3, improving the target positioning precision in the dense people stream, and finally constructing a multi-pedestrian detection model.

The invention belongs to the field of computer vision and artificial intelligence security, and discloses a self-adaptive frequency domain adversarial training method and device for a target detector, the target detector comprises a repair module and a pedestrian detector which are connected in series, and the input of the repair module is connected with the output of a patch detector; the self-adaptive frequency domain adversarial training method comprises the following steps: losses in joint training comprise standard target detection losses, repair consistency losses on a frequency domain based on a frequency domain image corresponding to a training image and a clean image, and repair dependence losses based on a detected average precision mean value; according to the invention, the end-to-end joint training is carried out through the restoration module and the subsequent pedestrian detector, and the optimization target of the restoration module is directly aligned with the improvement of the detection robustness, so that the confrontation disturbance is eliminated as far as possible, and meanwhile, the key semantic information of the detection task is reserved to the maximum extent. The separation of the performance of the repair module and the pedestrian detector is avoided, and the detection robustness is improved.

The invention provides a target detection method and system and a storage medium, and relates to the field of object detection, and the method comprises the steps: obtaining image data and laser radar scanning data corresponding to a current scene; calling a first pedestrian detection model based on an image to carry out pedestrian detection on the image data to obtain position information of a pedestrian target; calling a second pedestrian detection model based on point cloud data to carry out pedestrian detection on the laser radar scanning data to obtain pose information of a pedestrian target; and calling an external parameter matrix to match the position information and the pose information, and outputting three-dimensional position information of the pedestrian target in a space coordinate system. According to the invention, the detection and positioning of the pedestrian target can be realized without using a high-cost depth camera and a three-dimensional laser radar, and the detection precision of the pedestrian target is improved through two different modal data.

The invention provides a subway platform edge behavior abnormity real-time identification and early warning method based on deep learning, and relates to the technical field of deep learning, which comprises the steps of obtaining real-time video data, performing pedestrian detection and track extraction, calculating behavior characteristic parameters and a dynamic distance threshold value, performing danger grade division on pedestrians, and performing early warning and early warning. According to the method, time sequence feature extraction is carried out on potential dangerous targets, a behavior abnormity scoring mechanism is established, timely early warning of abnormal behaviors is realized, the accuracy and real-time performance of safety monitoring of the subway platform can be effectively improved, and the occurrence rate of safety accidents can be reduced.

The invention relates to the technical field of automatic driving, in particular to an automatic driving system based on pedestrian motivation and a control method, and the system comprises a data acquisition module, a pedestrian detection module, a pedestrian intention reasoning module and a vehicle control module. The data acquisition module is used for acquiring a video sequence and environment information of a traffic intersection when the intelligent driving vehicle drives to the intersection without the signal lamp. The pedestrian detection module detects pedestrians through video processing and a pedestrian recognition strategy and outputs pedestrian inter-frame images. The pedestrian intention reasoning module classifies pedestrian crossing motivations based on the trajectory and attitude information of pedestrians. The vehicle control module dynamically adjusts the vehicle speed according to the pedestrian motivation type and automatically brakes when necessary so as to ensure that pedestrians safely pass through the intersection. According to the system, self-adaptive comity of the intelligent driving vehicle to pedestrians at the intersection without signal lamps is realized, and the traffic safety and the adaptive capacity of intelligent driving are effectively improved.

The invention provides a single-mirror pedestrian height change perception and adaptive state perception Re-ID tracking method and system. The method comprises the following steps: acquiring a video stream and extracting a video frame; pedestrian detection is carried out on each video frame, and corresponding pedestrian features are extracted according to a pedestrian detection result; judging the confidence coefficient of each detection frame, and constructing a track instance when the confidence coefficient of each detection frame is greater than a threshold value; calculating an adaptive state sensing Re-ID cost matrix and a height change weighted intersection-to-parallel ratio cost matrix to obtain a final cost matrix; performing Hungary matching on the cost matrix to search the most matched trajectory instance for each pedestrian detection object; and integrating the track state, the track updating state and the detection frame state of the current time frame, and updating the track attribute. According to the method, hidden constraints of height changes of the detection frame and the tracking frame are considered, weights of the height changes are added on the basis of area constraints, and the problem of identity switching when pedestrians are shielded can be better solved.

The invention relates to the technical field of computers, and provides a cloud-edge collaborative abnormal behavior character recognition method, device and system, and the method comprises the steps: edge end equipment employs a trained pedestrian detection model to cut out a video sequence containing pedestrians from a received video sequence, and uploads the video sequence containing the pedestrians to a cloud server; the cloud server cuts a video sequence containing pedestrians according to a set time length to obtain video slices, the video slices and prompt words are input into a trained multi-mode large model to obtain an identification result, the prompt words are described in a video question and answer mode, option question and answer pairs in video question and answer are constructed according to related task types, and the identification result is obtained. The options in the video questions and answers are possible behavior types. The lightweight pedestrian detection is integrated on the side end equipment to reduce the transmission, reduce the processing data volume of the cloud server, solve the problem of high demand of cloud computing resources, and improve the efficiency and response speed of the system.

The invention relates to the technical field of cableway platform intelligent management, in particular to a cableway platform tourist falling detection method, system and device based on visual identification and a medium, and the method comprises the steps: collecting a video stream of a platform region, carrying out the pedestrian detection, obtaining the position of a bounding box of a tourist, and storing the position of the bounding box; a unique track ID and continuous track data are generated for each tourist by using a multi-target tracking algorithm; for each tourist, extracting a human body posture key point coordinate in each frame of the video stream and calculating a posture feature, and when the posture feature meets a preset posture collapse judgment rule, judging that the posture collapse is suspected; carrying out time sequence fusion on the continuous tracks of each tourist, and carrying out real falling event judgment based on the judgment condition of suspected posture collapse; and matching a preset platform area according to the position of the real falling event, outputting a risk level and triggering a corresponding control instruction. According to the invention, the falling behavior can be accurately and stably identified, and the safety protection capability of the cableway platform is improved.

The invention discloses a vehicle trunk automatic opening method and device and a computer program product, and the method comprises the steps: firstly, after the device fingerprint verification of a target vehicle is passed, obtaining N paths of surrounding video streams through a whole vehicle panoramic camera composed of N vehicle-mounted cameras; and then pedestrian detection and cross-shot pedestrian tracking processing are carried out on each frame of image in the N paths of video streams to obtain a pedestrian detection result and a pedestrian tracking result. Generating a BEV pedestrian trajectory according to a pedestrian tracking result; a target pedestrian BEV track is screened out from the target pedestrian BEV track; and after the identity verification of the target pedestrian is passed, obtaining a state diagram of the target pedestrian according to a pedestrian detection result, inputting the BEV track of the target pedestrian, the state diagram of the target pedestrian and the scene text information of the landmark building within the preset range around the target vehicle into a VLM in combination with prompt, and when the VLM judges that the target pedestrian has the intention of opening the trunk, opening the trunk. And the trunk is automatically opened, so that the safety of the target vehicle and the vehicle use experience of the target pedestrian are improved.

The invention discloses a vehicle and pedestrian detection method based on improved YOLOv8, and the method comprises the following steps: introducing a self-adaptive SE attention mechanism into a C2f module, and enhancing the extraction of key features; a double-path feature separation and fusion strategy is introduced into the C2f module, so that the detection precision of the model is improved; dynamic convolution optimization feature extraction and lightweight processing are introduced into the Bottleneck module, and global information collection is enhanced. According to the invention, by introducing an adaptive SE attention mechanism, a dual-path feature separation and fusion strategy and dynamic convolution optimization feature extraction and lightweight processing, the detection performance and calculation efficiency of the model are significantly improved; the method has remarkable advantages in the aspects of precision, calculation efficiency and model lightweight, and has wide application prospects in the fields of intelligent traffic monitoring, automatic driving systems, public safety and the like.

The invention provides a multi-view pedestrian detection method and system based on dual adaptive feature fusion, and belongs to the technical field of computer vision. According to the method, in a unified aerial view space, firstly, a cross-view feature selection module is introduced, and a multi-head self-attention mechanism is utilized to perform dynamic complementation and weighted fusion on features from different cameras so as to enhance the semantic integrity of a shielded area; then, a view channel map attention module is introduced, fused features are constructed into a view-channel map structure, modeling is carried out on the joint dependency relationship between view channels through a map attention perception multilayer perceptron combining global and local context pooling, self-adaptive reweighting of channel importance is achieved, and the self-adaptive reweighting of channel importance is realized; therefore, the discrimination of the fusion features is improved. And the finally obtained unified features are used for ground pedestrian positioning. According to the method, the detection performance bottleneck caused by visual angle shielding and channel redundancy is effectively relieved, and the quality of the ground occupancy map is remarkably improved.

The invention belongs to the field of computer vision and cross-modal retrieval, and particularly relates to an image-text pedestrian retrieval method based on image block replacement and cross-modal identity alignment, which comprises the following steps: acquiring a public data set of image and text description, and constructing a pedestrian re-recognition model PRCIA; inputting the data set into a PRCIA model for training and verification, and performing iterative updating on a training weight file through forward and backward propagation to obtain a trained PRCIA model; constructing a reasoning stage model, reserving double encoders and fusing global features in a reasoning stage, and ensuring the calculation efficiency; and inputting the test set into the reasoning stage model to obtain a detection result, thereby realizing text-based pedestrian detection. According to the method, fine-grained association between the image blocks and the text phrases is established through the PR module, cross-modal identity feature expression is enhanced through the CIA module, the accuracy and robustness of text-to-image pedestrian retrieval are remarkably improved, and the requirements of actual scenes are more easily met.

The invention discloses a pedestrian cross-border head tracking method and system based on multi-modal dynamic feature fusion, belongs to the technical field of computer vision, and aims to solve the technical problem of how to improve the accuracy and efficiency of pedestrian detection and the stability of cross-border head tracking. By adopting an advanced backbone network, a self-adaptive anchor frame mechanism, cross-scale feature fusion and other technologies, the detection capability of small targets and high-shielding-rate pedestrian targets is improved; an intelligent caching strategy and a model fusion method are adopted, so that repeated calculation of instance segmentation is reduced, and the efficiency of dynamic pedestrian segmentation is improved; through combination of multiple feature extraction and enhancement technologies, including optimized Mama, Transform and multi-granularity feature fusion, the purity and discrimination of the features and the adaptability to shielding and view angle changes are enhanced.

The invention discloses an improved PointPill three-dimensional target detection method based on fusion sparse enhancement and geometric attention, and belongs to the field of deep learning and three-dimensional target detection. According to the method, an SP-Point Pill improved framework fusing a sparse Pill enhancement mechanism (SEPA) and a point awareness space attention mechanism (PASA) is provided. A sparse Pilllar enhancement mechanism is fused, and geometric information of sparse Pilllar is efficiently filled through neighborhood real point borrowing and a virtual point interpolation strategy; the point sensing space attention mechanism generates a dynamic weight based on the point cloud density and the average distance, and preferentially enhances target related features; the pedestrian detection method and the pedestrian detection device cooperate with each other to form a link closed loop of information completion and feature focusing, so that the pedestrian detection precision is remarkably improved on a KITTI data set, and meanwhile, the efficient reasoning speed is maintained.

The various embodiments described herein include methods, devices, and systems for providing event alerts. In one aspect, a method includes obtaining a video feed. A frame of the video feed is analyzed at a first resolution to determine whether the frame includes a potential instance of a person. In accordance with the determination that the image includes the potential instance, a region is denoted around the potential instance. The region is analyzed at a second resolution, greater than the first resolution. In accordance with a determination that the region includes the instance of the person. a determination that the frame includes the person is made. An indication of the determination is stored for use in subsequent alert notification processing.

The invention discloses a night pedestrian detection method based on an artificial intelligence technology, and relates to the technical field of target detection, and the method mainly comprises the steps: preparing a night road pedestrian data set in advance, and carrying out the preprocessing; the method comprises the following steps: constructing a night pedestrian detection model NE-RTDETR based on RT-DETR neck enhancement, wherein the model is composed of a backbone network ResNet18, a neck enhancement network and a decoder; the data set is input into a NE-RTDETR model for training, and a trained best weight file is obtained; inputting a to-be-detected night road pedestrian picture into the NE-RTDETR model according to the trained weight to obtain information about whether a pedestrian exists in the night road pedestrian picture or not; and finally outputting a detection result and evaluating model performance. Through the steps, the problems that an existing detection method is insufficient in accuracy, high in missing detection and false detection rate and the like under the conditions of low illumination and fuzzy shielding at night are solved.

The invention relates to the technical field of computer vision and artificial intelligence, and discloses a public area pedestrian trajectory monitoring system based on cross-camera tracking and hybrid simulation completion. The system comprises a camera array module, a pedestrian detection and cross-camera tracking module, a monitoring blind area information base module, a hybrid drive pedestrian simulation engine module, a track fusion and complementation module and a simulation integration propulsion module. According to the invention, monitoring blind area tracks can be effectively filled, track continuity and prediction accuracy are improved, existing hardware is fully utilized to reduce cost, and modular design is easy to integrate and convenient to use. The generated complete track can be widely applied to scenes such as public safety monitoring, passenger flow analysis and emergency evacuation plan making in public places such as ports, subway hubs, cruise ship wharfs and commercial complexes, and supports a real-time monitoring or post-analysis mode.

The invention provides a multi-level domain adaptive pedestrian detection method and system based on a teacher model, and the method comprises the steps: training a target detector through employing source domain perspective image data with a pedestrian label, and obtaining an initial target detector; constructing a target domain teacher model and a cross-domain student model; and performing multi-level domain adaptive adversarial training by using the cross-domain student model, updating parameters of the target domain teacher model through index moving average of the weight of the cross-domain student model, generating a pseudo label to supervise the multi-level domain adaptive adversarial training of the cross-domain student model, and optimizing the cross-domain student model in combination with a designed loss function. Obtaining a final cross-domain student model; and detecting pedestrians in the overlooking fisheye image to be detected based on the final cross-domain student model. According to the technical scheme of the invention, the requirements of people flow statistics, safety monitoring and people flow guidance on rapid and accurate detection of pedestrians in a scene can be met.

The invention discloses a dense pedestrian detection method based on improved YOLOv11, and the method comprises the following steps: 1), judging whether an input image is a low-light scene or not, and carrying out the low-light enhancement preprocessing of the input image if the input image is the low-light scene; 2) replacing a standard convolution module in the YOLOv11 with a DBRF, wherein the DBRF is connected through convolution and residual errors of different scales; (3) a GETA is adopted to replace a C2PSA module in the YOLOv11; (4) reconstructing a Neck network structure of the YOLOv11; and 5) outputting a pedestrian detection result in combination with the detection head. Multi-scale pedestrian features are extracted by using DBRF, occlusion pedestrian feature expression is enhanced by using GETA, and a small target detection layer is added in a Neck part for improving small target detection capability, so that the accuracy and recall rate of pedestrian detection in a dense crowd scene are effectively improved, the omission ratio of small targets is reduced, and the limitation of a traditional technology is effectively overcome. And higher application requirements are met.

The invention discloses a pedestrian detection method and system fusing multi-branch feature extraction and global context enhancement, and the method specifically comprises the steps: obtaining an urban scene pedestrian image data set, carrying out the preprocessing, and dividing the data set into a training set and a verification set; an improved YOLOv12 network architecture is constructed, the improved YOLOv12 network architecture comprises a backbone network Backbone, a neck network Neck and a detection head Head, an improved C3K2 module is adopted as a feature extraction core unit of the backbone network, and an effective feature layer is extracted; a DySample dynamic up-sampling module is adopted, the multi-scale feature fusion effect is optimized, and the extracted features are refined; a global context enhancement module GCEM is additionally arranged between the Neck and the Head, and pedestrian global feature expression is enhanced; and outputting the category and location of the target according to the results of the classification branch and the regression branch. The pedestrian detection accuracy and robustness in a complex scene are improved, high reasoning efficiency is kept, and the method is suitable for real-time detection scenes such as intelligent traffic and security monitoring.

A work vehicle includes an operator seat and a seat sensor configured to provide a first signal in response to presence of an operator seated in the operator seat. A pedestrian detection system includes an electronic controller operatively coupled with one or more sensors. The pedestrian detection system is configured to detect presence of a pedestrian in proximity to the work vehicle. A controllable switch device operatively coupled with the seat sensor and the pedestrian detection system. The controllable switch device is configured to provide a second signal to activate the pedestrian detection system in response to the first signal provided by the seat sensor.

The invention discloses a pedestrian detection method and device in a complex environment, equipment and a storage medium, and relates to the technical field of target detection, and the method comprises the steps: receiving to-be-detected pedestrian image data, carrying out the feature extraction of the to-be-detected pedestrian image data through a self-adaptive multi-scale convolution operation, and obtaining the feature of the to-be-detected pedestrian image data; enhancing key region features in the primary feature map according to a large separable kernel attention mechanism, generating a fusion feature map, performing processing on the fusion feature map according to content perception feature recombination up-sampling, generating a high-resolution feature map, and performing pedestrian detection according to the high-resolution feature map. And obtaining a detection result. According to the invention, the accuracy of pedestrian detection can be improved.

A system for alerting pedestrians to a vehicle includes a sensing system having at least one imaging sensor in communication with a controller. The controller includes a processor and a memory. The memory stores a pedestrian detection module and a pedestrian alert module. The pedestrian detection module including instructions configured to cause the processor to identify a pedestrian in an output of the at least one imaging sensor using image analysis and instructions configured to cause the processor to respond to an identified pedestrian by determining a reduced sensory state of the pedestrian. The system generates an ambient environment model based on an output of the at least one imaging sensor and a connection to at least one external data source. The pedestrian alert module causes the processor to generate a pedestrian alert that contrasts with the ambient environment model. An external alert system implements the pedestrian alert.

The utility model discloses passing pedestrian detection and early warning complete equipment, which comprises a pedestrian detection unit, an early warning trigger unit and an early warning prompt unit which are arranged by adopting a modular structure, a sensor is arranged in the pedestrian detection unit and is used for detecting the current environment condition, and the current environment condition comprises whether a person and / or an animal is going to pass through an intersection or not. Or whether pedestrians, animals, obstacles or collapses exist on the road or not; a logic control circuit is arranged in the early warning trigger unit and is used for judging the danger level of driving according to the current environment condition detected by the sensor, setting three output interfaces for the speed value, the number and the volume value, and respectively judging whether to send out an early warning signal or not; and the early warning prompt unit sets three modes for picture brightness, flicker frequency and prompt tone frequency, receives a control signal of the early warning trigger unit, and prompts vehicles or pedestrians in different modes according to danger levels.

The application discloses a kind of pedestrian action continuous detection identification method, device, storage medium and equipment, belong to action detection identification field.It includes: the video to be detected is divided into multiple video segments and selects key frame;Frame image of video segment is input into pedestrian detection model, and detection frame is obtained on each frame image;The detection frame of the latter frame image is associated with the detection frame of the former frame image, and the detection frame of the two frame images on association is set with the same number;All frame images of video segment and its detection frame are input into action recognition model, and the action category of each detection frame of key frame is obtained;The action category of the detection frame of non-key frame is set as the action category of the detection frame of key frame with the same number thereof.The application realizes the frame-by-frame continuous detection identification of human action in video, solves the problem that only relying on key frame detection brings deviation to overall detection evaluation of video, and improves the visual experience of video action detection identification presentation.

The invention relates to an unmanned aerial vehicle detection-based method and system for detecting pedestrians who do not give way to a vehicle, and the method comprises the steps: carrying out the image collection through an image collection device carried by an unmanned aerial vehicle, and obtaining a traffic image sequence; performing multi-target detection on all moving targets in the frame image through a target detection model to obtain a target detection frame and corresponding confidence; performing trajectory tracking processing on the target detection frames and the confidence coefficients of all the frame images through a target tracking algorithm to obtain a tracking trajectory of the moving target; obtaining a zebra crossing control area, calculating a pedestrian warning point and a vehicle warning point based on the tracking trajectory when the zebra crossing control area in the frame image is detected to have the pedestrian trajectory and the vehicle trajectory, and calculating a vehicle dangerous area according to the vehicle warning point; and according to the tracking trajectory, the pedestrian warning point and the vehicle dangerous area, analyzing a position relationship between the pedestrian and the vehicle, and according to the position relationship and a preset non-comity pedestrian judgment logic, determining whether there is a behavior that the vehicle does not comity the pedestrian.

The application discloses a dizziness lamp control method and system based on human body depth estimation and target detection, comprising: acquiring a first data set and a second data set; training a pre-constructed depth estimation model according to the first data set, and then training a pre-constructed pedestrian detection model according to the second data set to obtain a human body depth estimation network; collecting a current monocular image, inputting the current monocular image into the human body depth estimation network to obtain human body depth information; and adjusting the brightness of the dizziness lamp according to the human body depth information. The application combines target detection and depth estimation to obtain human body depth information, can automatically and dynamically adjust the brightness of the dizziness lamp according to different target pedestrian distances, improves the single use time length of the dizziness lamp under the premise of ensuring the effect of the dizziness lamp, and can be widely applied to the technical field of light control.

The invention relates to a blind area pedestrian detection method and device and electronic equipment, and relates to the technical field of blind area detection. The blind area pedestrian detection method comprises the following steps: constructing at least two tracking queues and a blind area queue; traversing the tracking information of the pedestrians in the at least two tracking queues; under the condition that the traversed pedestrians are the first type of pedestrians, updating the blind area queue and the first tracking queue according to a matching result between the tracking information of the first type of pedestrians and the tracking information of blind area pedestrians in the blind area queue; under the condition that the traversed pedestrians are the second type of pedestrians, updating a blind area queue and a second tracking queue according to the lost frame number of the tracking information of the second type of pedestrians in the at least two tracking queues; and outputting detection information of pedestrians in the blind area according to the blind area queue and the at least two tracking queues. According to the invention, on the premise of not deploying a shooting assembly or other sensors for collecting images in the blind area, effective detection and tracking of pedestrians in the blind area are realized, so that the hardware cost of the system is directly reduced.

The invention, which relates to the technical field of target detection, discloses a target detection method for reducing a vehicle collision risk, comprising: acquiring real-time image data of an environment in front of a vehicle; inputting the real-time image data into an optimized target detection neural network model, and identifying a pedestrian target in the image and position information of the pedestrian target; wherein the neural network model is based on a YOLO architecture, and optimization is carried out for a pedestrian detection task; a collision risk is calculated based on the identified pedestrian location information, and an active safety system of the vehicle is triggered when the risk exceeds a threshold. According to the scheme, the detection precision and real-time performance of the intelligent driving vehicle on pedestrians, especially small-scale pedestrians in a complex traffic environment can be remarkably improved, the collision risk caused by missing detection or false detection is effectively reduced, and the active safety performance of the vehicle is improved.

This invention discloses a real-time pedestrian intrusion detection method and system based on machine learning for highways, belonging to the field of pedestrian detection technology. It includes a device deployment module for acquiring key data information; a data processing and storage module for storing and preprocessing the acquired key data information; an analysis module, including a confidence assessment unit, a perception module analysis unit, and a multi-source dynamic fusion unit, for determining whether pedestrian intrusion occurs; and an alarm module that generates a corresponding alarm upon determining the presence of a pedestrian intrusion. This invention can more accurately and quickly identify pedestrian intrusions through comprehensive analysis of multi-source data. Furthermore, it can assign corresponding weights to different perception modules based on different weather conditions, enabling the system to autonomously adapt to perception characteristics under extreme weather conditions, thereby reducing environmental interference and improving detection accuracy.