34 results about "Person detection" patented technology

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 image recognition device (10) includes: a person detection unit (14) that detects people in a captured image and determines the detection position and detection size of the people in the captured image; a frame determination unit (16) that determines the size of a frame that is different from the detection size based on the determined detection position of the people; an image generation unit (18) that generates a display image by superimposing the frame of the determined size onto the detection position of the captured image; and a display control unit (20) that causes the display device (28) to display the generated display image.

This invention discloses a scale-redistribution-based top-view fisheye pedestrian detection method, belonging to the field of target detection technology. In the feature generation stage, this invention employs frequency-preserving downsampling, concatenating low-frequency contours and high-frequency textures after sub-pixel rearrangement and wavelet decomposition in the channel to achieve spatial depth swapping. In the feature fusion stage, it increases the participation ratio of high-resolution features through adaptive upsampling for geometric alignment. In the spatial computation stage, it divides features into high-response and low-response regions by learning a spatial weight map, performing focused convolution only on high-response regions. In the optimization stage, it constructs a joint regression mechanism, jointly modeling overlap consistency and distribution distance, adjusting sample gradient contributions based on detection difficulty, and controlling the angle supervision intensity according to the target aspect ratio, forming a scale redistribution that spans multiple stages. This invention achieves stable detection of small-scale and distorted targets in top-view fisheye images, improving the detection consistency between edge and dense regions.

The application provides a shore bridge ship side anti-punching person detection method and system based on machine vision, wherein the method comprises: obtaining a detection frame of a target by using machine vision technology, and transforming a personnel detection frame of a second camera into an image coordinate system of a first camera; calculating an estimated pixel coordinate of a spreader when the spreader is lowered to a personnel height in real time according to a position, a size of a current spreader detection frame and a distance from the spreader to the camera; determining a distance between the spreader and the personnel by using the estimated spreader and the personnel detection frame pixel coordinates; and generating an alarm information when the distance between the spreader and the personnel is less than a safety distance. The application calculates the distance from the personnel and the spreader to the camera by using a depth estimation network and a camera imaging principle, estimates the distance between the spreader and the personnel when the spreader is lowered to the personnel height according to the distance, and generates an alarm information according to the distance value, thereby solving the problem that the automatic spreader anti-punching person alarm judgment cannot be performed due to the change of the height from the ship to the camera.

Labeling machine (1) for containers, of the rotary type and with lateral docking stations (3A to 3C) for labeling units (4A to 4C), and with a common operating unit (5) for the docking stations (3A to 3C), wherein the operating unit (5) can be pivoted and / or moved in a circumferential direction (6) into different access areas (7AL to 7CR) to the docking stations (3A to 3C) and labeling units (4A to 4C), characterized by an optoelectronic monitoring device (9) for person detection in the access areas (7AL to 7CR); and a control (12) for automatically blocking a hub (13, 14) of the docking stations (3A to 3C) based on person detection in the access areas (7AL to 7CR) that are not directly visible from the control unit (5).

This invention provides a bidirectional feedback collaborative optimization method, medium, and device for pedestrian detection and tracking. The method includes: generating an initial set of detection boxes using a detector for the current frame image; generating virtual detection boxes using trajectory-guided detection completion; obtaining the detection box set after verification; and employing bidirectional feedback collaborative optimization of the detector and tracker, including: feeding the virtual detection boxes as positive hard examples back to the detector for optimization; feeding back high-confidence real detection boxes from the detection box set to the tracker for tracker updates; and performing a second data association using the verified detection box set and the tracker optimized by bidirectional feedback, ultimately outputting a set of pedestrian trajectories. This method can solve the problems of error accumulation, false detection propagation, and trajectory fragmentation caused by the unidirectional serial interaction between the detector and tracker in existing MOT systems.

The application discloses a personnel detection system and a personnel detection method, and belongs to the technical field of intrusion detection. The system comprises the following steps: a plurality of target areas have authorized first terminals establishing communication connections through short-distance wireless communication modules to form a distributed communication network; the first terminal is used for broadcasting a short-distance wireless communication signal and scanning the short-distance wireless communication signal of other terminals within a preset distance range, and the first terminal is also used for publishing the scanned short-distance wireless communication signal to the distributed communication network; the first terminal is also used for acquiring signal strength indication information of the target area based on the short-distance wireless communication signal in the distributed communication network, and determining un-authorized personnel detection information of the target area based on the signal strength indication information. The system can realize accurate detection of un-authorized personnel, has the advantages of simple deployment, low cost, strong environmental adaptability and high privacy security.

The utility model relates to a kind of arch top unmanned detection devices of power generation diversion tunnel inclined shaft section lining structure, and it is related to tunnel detection technical field, it includes traction mechanism, walking mechanism, adjusting mechanism and detection mechanism.Traction mechanism is connected with walking mechanism;Adjusting mechanism includes first telescopic unit and second telescopic unit, first telescopic unit is installed in walking mechanism and is connected with second telescopic unit;Detection mechanism includes detection platform, air coupled radar and ranging sensor, detection platform is installed in second telescopic unit, and air coupled radar and ranging sensor are all installed in detection platform;Second telescopic unit is connected with detection platform;First telescopic unit is used to drive second telescopic unit and detection platform reciprocating linear motion in first direction relative to walking mechanism;Second telescopic unit is used to drive detection platform to lift in second direction relative to walking mechanism;First direction and second direction have included angle.It is flexible to control, and stability is high, safety is high, and detection result is accurate and reliable.

The present application discloses an intelligent fire alarm control system and method, a device, and a readable storage medium. The control system comprises: a fire detection control module, which is used for detecting a fire signal, and outputting a first control signal to a controller when the fire signal is detected; a person detection control module, which is used for detecting a person signal, and outputting a second control signal to the controller when a person is detected; and the controller, wherein the controller is used for: when the first control signal and the second control signal are received at the same time, not outputting an alarm control signal; and when the first control signal is received while the second control signal is not received, if smoke and fire are artificially generated and a person has subsequently left, not outputting an alarm control signal, or otherwise, outputting an alarm control signal to give an alarm. The present application avoids the risk of false alarms, and improves the reliability and accuracy of fire alarms.

The present application relates to the technical field of intelligent detection, and discloses a seat person detection method based on a capacitive device, which comprises the following steps: collecting multiple initial capacitance values of a capacitive touch sensor of a target seat based on a fixed sampling period, and calculating an empty seat reference capacitance value; obtaining a current sampling time and an instantaneous capacitance value, and performing drift compensation on the empty seat reference capacitance value to obtain a current reference capacitance value; calculating a capacitance difference value of the two, and counting a super-threshold duration which is not lower than a preset capacitance trigger threshold; when the super-threshold duration meets the standard, determining that a user is seated in combination with a capacitance change rate; and when the super-threshold duration does not meet the standard or the capacitance difference value falls back to a preset steady-state threshold, determining that the user is getting off the seat. The present application further provides a seat person detection system based on a capacitive device, a device and a storage medium. The present application can improve the detection accuracy and stability of a seat state.

The present application relates to a kind of single-spectrum night pedestrian detection method and system based on foreground-background contrast attention.The method comprises: the image to be detected is input into the backbone network to extract different scale features containing different semantic information and detail information;Different scale features extracted by the backbone network are input into the feature fusion network, the different scale features are fused by the feature fusion network, and the feature in the foreground feature and background information two dimensions is adaptively adjusted by foreground-background contrast attention, i.e.strengthen the attention degree of foreground feature, weaken the attention degree of background information, so that the network can focus more on foreground feature in night scene;The multi-scale features after fusion are sent to the detection head, and the detection head maps the multi-scale features after fusion into a prediction box, so as to obtain pedestrian detection result.The present application can effectively improve the accuracy of night pedestrian detection.

To provide a worker detector, worker detection method, and program making it possible to alleviate a load on a monitoring person in inspection work of an electric power facility, and allowing a worker to safely carry out inspection work.SOLUTION: A detector 100 includes a determination unit 153 that determines whether a worker exists in either a detection area, which is defined in a fixed point image acquired by an acquisition unit 152, or a danger area located inside the detection area, and a rendering unit 154 that, when it is determined by the determination unit 153 that a worker exists in either the detection area or danger area, renders frames representing the detection area and danger area in the fixed point image acquired by the acquisition unit 152 and a mark indicating the position of the worker.SELECTED DRAWING: Figure 2

This invention relates to the field of data processing technology, specifically to a 3D binocular person attribute recognition and counting method and system integrating AI and REID. The system includes an image acquisition and person detection module, a feature extraction module, a feature enhancement module, a motion state matching module, a deduplication module, and a correction module. This invention solves the problems of low recognition accuracy and easy misjudgment in existing technologies when people have similar features or sudden changes in motion.

A technique for determining a presence of a person in a room may include an electronic device transmitting an electromagnetic wireless signal of a first sensor. The technique may include receiving an electromagnetic return signal from the electromagnetic wireless signal. The technique may include detecting a potential target in the room based on the electromagnetic return signal. The technique may include determining that the potential target is in the room using a second sensor. Responsive to determining the potential target is in the room, the technique may include saving a training signature of the electromagnetic return signal for training a machine learning model. This technique can be repeated to obtain a set of training signatures corresponding to potential targets. The technique may include training, using the set of training signatures, the machine learning model to detect when a target is in the room using the first sensor.

This invention proposes a method for constructing a personnel detection model based on global-local alignment and frequency domain adaptive fusion in fire smoke scenarios. First, near-infrared and thermal imaging dual-modal images are acquired, and multi-scale features are extracted using a dual-stream YOLOv11 backbone network. Second, a GLAFC fusion module is designed: first, near-infrared features are corrected using thermal imaging as a spatial reference; then, high-frequency textures from near-infrared imaging and low-frequency thermal radiation from thermal imaging are adaptively enhanced in the frequency domain; next, cross-modal repair and fusion are achieved through confidence-aware cross-attention. Then, the fused features are enhanced by SPPF and C2PSA and input into the neck network, where upsampling, stitching, and a C3k2 module are used to aggregate semantic and localization information to generate a multi-resolution detection map. Finally, a multi-task loss function with structural alignment consistency constraints is used to train the model. Experiments show that this invention can significantly reduce the false negative rate and improve detection accuracy in extremely dense smoke environments.

An image processing device of a person detection system mounted on a moving body is configured to: detect, in image data obtained from a camera, an area in which an obstacle appears; determine whether the area meets an upper body detection process condition that the obstacle in the area is distanced from a road surface within a predetermined range from the camera; perform an upper body detection process in which the area of the image data is compared with upper body comparison data to determine whether the obstacle in the area is a person, for the area that meets the upper body detection process condition; and perform a whole-body detection process in which the area of the image data is compared with whole-body comparison data to determine whether the obstacle in the area is a person, for the area that does not meet the upper body detection process condition.

This invention discloses an interaction category detection method, apparatus, device, and storage medium, relating to the field of image recognition technology. The method includes: inputting the image to be recognized into the object detection network of a pre-trained helmet-wearing detection model to obtain at least one person detection box, at least one helmet detection box, and at least one set of interaction information corresponding to interactive detection objects; inputting the person detection boxes, helmet detection boxes, and interaction information into the helmet-wearing detection model's wearing relationship determination network to determine a first person detection object with a wearing relationship to the helmet detection object and a second person detection object without a wearing relationship to the helmet detection object; and marking the first person detection object and the second person detection object in the image to be recognized. The technical solution of this invention can more reliably distinguish between people wearing helmets and those not wearing helmets, significantly reducing false detections.

This invention provides a channel-based terahertz security inspection data extraction and control method based on photoelectric switch control. It solves the problems of existing methods, such as their inability to handle large-volume, multi-person inspections, poor real-time performance, and slow data processing. By using four pairs of photoelectric switches, the security inspection channel is divided into three areas, and an independent data column is established for each area. Each data column can only contain the personal information code and corresponding security inspection data of one person being inspected at a time, thus enabling independent processing of data in different areas (i.e., multi-threaded operation) without interference. This allows for simultaneous terahertz security inspections of different targets located in different areas. On the one hand, it effectively improves the data extraction and processing effect and increases inspection efficiency, thus effectively handling large-volume inspection requirements. On the other hand, it effectively ensures the uniqueness of security inspection data, ensuring accurate and reliable security inspection results.

The application discloses a place illegal resident detection system based on power consumption behavior analysis, which comprises a data communication and transmission unit, a temperature and humidity data acquisition unit, a calculation and reasoning unit, a data storage unit and a power data high-frequency acquisition unit. The power data high-frequency acquisition unit has data acquisition capability for users, and satisfies a sampling rate of not less than 6.4KSPS. The application relates to the technical field of power consumption detection. The place illegal resident detection system based on power consumption behavior analysis comprises a device for illegal resident detection of nine small places. The device has integrated acquisition, calculation and reasoning capabilities, can satisfy the advantages of rapid deployment and rapid identification, and can satisfy the edge computing capability of the composition of the detection device, the hardware part and the system part. The detection device has high engineering capability, and the detection method of the detection device can satisfy the local computing capability, can continuously correct the accuracy of the model output, and has batch application.

This invention discloses a method, apparatus, device, and storage medium for blackboard writing behavior recognition. The method includes the following steps: acquiring video frames of a video to be recognized and the person detection results of the video frames; determining the face orientation of the person based on the person detection results; filtering out target persons that meet the orientation requirements based on the face orientation; detecting the distance between the target person and the blackboard; and determining that blackboard writing behavior exists when the distance between the target person and the blackboard is within a preset threshold range. Compared with traditional methods, this invention does not require pre-setting the specific posture of blackboard writing, reducing the workload in the early stages. It also avoids mistakenly identifying actions similar to blackboard writing as blackboard writing behavior, resulting in high accuracy in action recognition and low computational load.

This invention proposes a method, device, vehicle, and computer-readable storage medium for detecting occupants inside a vehicle. By installing pressure sensors on the seats inside the vehicle, the system can cover the possible locations where occupants may be present, avoiding blind spots in detection. Furthermore, by arranging the pressure sensors in an array, the target detection model comprehensively analyzes the data detected by the pressure sensors in different distributions to detect occupants. This allows the system to perceive changes in occupant body shape and posture, accurately distinguishing between occupants and heavy objects. Based on current environmental data, the system clarifies the characteristics of the detection environment and determines the target judgment threshold by combining the influence of the environment on pressure detection, thereby improving the accuracy of occupant detection.

The application discloses a visual language model guided indoor fire pedestrian detection and re-identification system and belongs to the technical field of computer vision and artificial intelligence. In view of the problems of instability in complex scenes, discontinuity in identification, lack of semantic guidance mechanism and cross-frame semantic consistency of the existing indoor fire pedestrian detection system, the application encodes the pedestrian-related text description through a text feature generation module, extracts multi-scale visual features through a visual feature extraction module, generates a semantic guidance weight matrix through a text and image feature fusion module, enhances the visual features and outputs a detection result through a semantic guidance detection module, realizes cross-frame identity consistency maintenance through a semantic guidance pedestrian re-identification module, and dynamically updates a pedestrian feature library through an online learning module. The system improves the detection precision and robustness under the environment of smoke, uneven illumination and flame interference, and enhances the continuity and stability of identity re-identification.

The present application relates to a radar sensor with a wide field of view. A millimeter wave radar sensor module is proposed with a specific architecture that enables a 180-degree field of view (FoV) to be achieved at an affordable system cost. The proposed millimeter wave radar sensor module is used in passenger cars, commercial vehicles, autonomous guided vehicles (AGV), as well as motorcycles and bicycles, thereby enabling a variety of application features: intelligent parking assistance, replacement for PDC sensors, determination of the distance, speed, and angle of targets, including their tracking, person detection, person tracking, tailgate and door protection by limit opening, gesture and kick sensing, and other applications. In the case of two- and three-wheeled vehicles, the same device can also cover blind spot detection applications in an affordable system manner.

This application provides a method for training a detection model, a pedestrian detection method, an apparatus, and related equipment, belonging to the field of computer vision technology. The method includes: acquiring sample pedestrian images and sample background images that match the scene in the sample pedestrian images; inputting the sample pedestrian images and sample background images into an initial detection model; extracting features from the background in the sample background images to obtain initial background features, and extracting features from the pedestrians in the sample pedestrian images to obtain initial pedestrian features; fusing the initial background features and initial pedestrian features to obtain sample fusion features; predicting the background in the sample fusion features to obtain sample background features; calculating a background loss value based on the sample background features and initial background features; training the initial detection model based on the background loss value to obtain the trained detection model. This application can improve the accuracy of pedestrian detection.

The present application relates to the technical field of artificial intelligence, and provides a personnel detection method and device, electronic equipment, storage medium and program product, which comprises the following steps: obtaining historical monitoring data of target personnel in a gridded detection space grid within a first preset total time length before the current time; taking a sampling time length as a sliding window, and converting the historical monitoring data of each grid into a plurality of monitoring data slices; constructing a target feature vector according to the monitoring data slices of all grids within a preset number of sampling time lengths before the current time; inputting the target feature vector into a personnel detection model, and obtaining a personnel existence detection result of the detection space at the current time. Based on artificial intelligence technology, in the scene where the current radar cannot detect personnel, the historical monitoring data of the personnel detected by the radar is extracted for features, so that whether the personnel is in the current detection space can be accurately predicted according to the personnel's line of movement, and the control demand of the personnel in a specific scene can be met.

A system, method and apparatus for detecting the presence of a person in an area monitored by an electronic person detector that may operate both as a security motion sensor and a home automation occupancy sensor.