433 results about "Person detection" patented technology

Embodiments of the present invention provide a method, system and computer program product for bot detection for network distributable markup. In accordance with an embodiment of the present invention, a page request for distributed markup can be processed to incorporate embedded fragment within the requested page. For instance, the fragment can include a script enabled to detect human activity within the requested page such as a mouse movement. Alternatively, the fragment can include an extraneous markup artifact. The requested page subsequently can be returned to the requestor and the embedded fragment can be monitored to detect the presence of a bot depending upon the activation of the artifact. For example, where human activity can be detected within the page or where the extraneous markup artifact becomes activated despite the extraneous nature of the artifact, a human requestor can be concluded. However, where no human activity is detected in the requested page, or where the extraneous markup artifact remains unactivated, a bot requestor can be determined.

In masquerading determination processing, a masquerading determination unit reads image data representing an image of an identification target on which a striped pattern is projected from an image storage unit to extract the striped pattern appearing in a face region of the image represented by the read image data. Subsequently, the masquerading determination unit determines whether a stripe in the face region in the image is a straight line or not. When the stripe is a straight line, because the identification target is a plane object such as a photograph or an image display device so that it can be determined that the target is at least not a person himself, the masquerading determination unit determines that the target masquerades. On the other hand, unless the stripe is a straight line, because the identification target has a solid configuration having three-dimensional irregularities to have a possibility of being a person himself, the unit determines that the target might not masquerade.

The invention discloses a pedestrian detection method based on a deep learning technology. The method comprises the steps that firstly, a binary classification model is trained through a step-by-step migration strategy on the basis of transfer learning to initialize final model parameters; secondly, pedestrian detection work is completed by adopting and modifying a currently popular and efficient Faster RCNN frame, and on the basis of the CNN characteristics of the frame, not only can images with any scales be processed, but also the detection speed is high. Compared with the prior art, the method has the advantages that the network does not need to be specially designed, existing available data is fully utilized, a good experiment result still can be achieved by adopting a general network structure, the advantages of a deep convolution network are fully achieved, and the advantages of being simple in design, good in robustness, high in detection accuracy and low in omission ratio are achieved.

The invention discloses a pedestrian detection method which comprises the following steps of: obtaining a candidate detection window in the current-frame image by a human body detection method based on HOG characteristics; then determining the corresponding relationship between the current candidate detection window and the human body detection window in the previous-frame image, and standardizing the size of the current candidate detection window; determining to adopt an image division method based on shape prior or an image division method based on the combination of motion information and shape prior for the current standardized candidate detection window according to whether correspondence is established so as to obtain an object division mask of the current standardized candidate detection window; and finally, verifying whether the current candidate detection window is detected wrongly by a pedestrian classifier jointly trained by the object division mask and HOG detection score. Since the motion information and shape prior are integrated into the image division method, the accuracy of object division is improved, and the detection judgment of the candidate detection window is further improved, thus the error detection rate is effectively reduced, and the detection accuracy is improved.

A projection type video display apparatus in which a display device transmits light emitted from a light source to project a video written into the display device is characterized by comprising person detection means for detecting a person who exists within a projected video area, and control means for changing the state of the projected video when the person detection means detects the person.

Provided is a suspicious person detection method. First, a normal similar facial image search is carried out. Next, facial images, which are detected automatically from the input images and specified manually, are specified to be determined. Next, similar faces are searched for limited time on a time axis on the database. Next, the number of search results that distance between the features is lower than predetermined value is calculated and it is determined that the number of appearances is large and a possibility of a prowling person is high if the number of cases is large, and otherwise a possibility of prowling person is low. Last, a similarity between a facial image of a pre-registered residents and a facial image of a person whose number is large is calculated, and it is re-determined that the person is residents if the similarity is high, regardless of the determination.

A photographing device, which includes a photographing section; a display section that displays a photography image taken by the photographing section; a person indicator setting section that sets a person indicator prepared beforehand on the display screen; an operation section that operates a target position and / or a target size of the person indicator; a person detection section that detects a position and / or a size of a person in the photography image; a comparison section that compares the detected position and / or size of the person in the photography image with the target position and / or size; a guiding instruction section that generates an instruction for the person to be photographed to guide toward a direction coinciding with the compared result; and a control section that controls the photographing section such that, when a predetermined condition is satisfied after the compared result falls within a predetermined allowable error, a shutter is activated.

A face recognition system based on adaptive learning includes a specific person detection and tracking unit for detecting and tracking a specific person from a moving image. A facial feature extraction unit extracts a plurality of facial feature vectors from the detected and tracked specific person. A face recognition unit searches for a given registration model by comparing the extracted facial feature vectors with facial feature vectors of the registration models previously stored in a user registration model database. A learning target selection unit selects a facial feature vector to be added to a record of the given registration model from among the extracted facial feature vectors. A registration model learning unit adds and updates the selected facial feature vector to the record of the given registration model.

An image processing system processes a sequence of images to generate a statistical model for each of a number of different persons to be tagged so as to be identifiable in subsequent images. The statistical model for a given tagged person incorporates at least one appearance feature, such as color, texture, etc., and at least one geometric feature, such as shape or position of a designated region of similar appearance within one or more images. The models are applied to subsequent images in order to perform a person detection, person location and / or person tracking operation. An action of the image processing system is controlled based on a result of the operation.

This invention discloses a living person detection method applied to face recognition. The method includes the following steps that: modeling is performed on the background of a current recognition environment; face detection is performed: the locations and sizes of faces are obtained and are adopted as the initial position of next face detection; the similarity between the background of inputted face images and a modeling background image is calculated through adopting an SSIM method; when the faces are detected, ultrasonic waves are utilized to measure the distance between a camera and an obstacle, and a classification probability value f is obtained through utilizing a logistic regression model; and an SSIM value and the value f are combined linearly, so that a final result R can be obtained, and when R is smaller than a certain threshold value, a spoofing attack is determined. The living person detection method applied to face recognition of the invention has the advantages of high success rate and excellent stability.

A person detection apparatus and method, and a privacy protection system using the method and apparatus, the person detection apparatus includes: a motion region detection unit, which detects a motion region from a current frame image using motion information between frames; and a person detecting / tracking unit, which detects a person in the detected motion region using shape information of persons, and performs a tracking process on a motion region detected as the person in a previous frame image within a predetermined tracking region.

A behavior analysis / monitoring device includes: a person detection unit configured to detect a person(s) from image information obtained by capturing images covering an area around an item placement area; a part-of-interest detection unit configured to detect, for each person detected by the person detection unit, a part of interest set in a part of an upper body of the person excluding hands and arms; a position measurement unit configured to measure a position of the part of interest detected by the part-of-interest detection unit; and an item pick-up action determination unit configured to obtain a displacement of the part of interest based on the position of the part of interest obtained by the position measurement unit and to determine whether each person detected by the person detection unit performed an item pick-up action based on the displacement of the part of interest of the person.

The invention discloses a night pedestrian detection method based on statistical features of infrared pedestrian brightness. According to the method, gray average information of parts of one pedestrian and a negative sample in a sample database is firstly subjected to statistical processing, a mapping interval boundary is determined by the aid of the information, and a brightness histogram feature for distinguishing vote interval division is constructed; then, a histogram feature in the gradient direction is calculated, and the two features are combined to form a final feature descriptor; secondarily, model training is performed by the aid of Adaboost in combination of a decision tree, and pedestrian determination and positioning are performed with a sliding window scanning method; finally, when a classifier obtains lower degree of confidence through classification judgment on a certain detection frame, a brightness interval template is adopted for secondary judgment, so that night pedestrian detection is realized. Pedestrians in the night environment are effectively detected, and the method has the characteristics of high detection rate and adaptability.

A person detection apparatus controls activation of a crash absorption device such as an external airbag of a vehicle, whereby when the vehicle is about to collide with a detected object, a judgement is made as to whether the object is a person. If the object is judged to be other than a person, activation of the crash absorption device is inhibited. It is rendered easier or more difficult for an object to be judged as being a person, in accordance with whether or not the vehicle is operating in an environment with a relatively high probability that persons will be present.

A person detection unit in an image processing apparatus is input an image captured by an image capturing unit and detects a person from the input image. An observation target determination unit determines an observation target in the image captured by the image capturing unit according to an action taken by the person detected by the person detection unit.

The invention relates to a person identification method and a person identification device, and belongs to the technical field of person detection and identification. The method comprises the steps of identifying face features of a target person based on sample face features in a face database, identifying clothes features of the target person according to sample person clothes features in a clothes database if judging that a first identification confidence with the highest degree of face matching is smaller than a preset first threshold, acquiring a second identification confidence with the highest degree of clothes matching, and identifying the target person according to the first identification confidence and the second identification confidence. According to the invention, when the identity of a person is hard to identity accurately based on the face features, the identity of the person is identified based on both the face features and the clothes features. Therefore, the accuracy of person identification is improved, the diversity of person identification object samples is satisfied, and person identification can be handled more flexibly.

The invention discloses a video-based multi-target pedestrian detection and tracking method. A YOLO3 target detection algorithm with good speed and accuracy is utilized, video images under different scenes are constructed, and a detection model is trained, so that the influence of illumination change and visual angle change is overcome, and efficient detection of multi-target pedestrians is ensured; the multi-target pedestrian is effectively tracked based on the Kalman filtering algorithm and the Hungary algorithm effectively tracks, the problem that repeated target detection often exists in multi-target detection is avoided, thereby realizing the multi-target pedestrian tracking method taking the Deep-SORT algorithm as the core. The method has the characteristics of efficient multi-targetpedestrian detection and efficient multi-target pedestrian tracking.

The present invention provides a room information inferring apparatus, a room information inferring method, and an air conditioning apparatus. The information regarding a room, such as the shape of the room, is inferred simply and precisely. A room information inferring apparatus for inferring information regarding a room includes: an imaging unit that captures an image of a room that is to be subjected to inferring; a person detection unit that detects a person in an image captured by the imaging unit, and acquires the position of the person in the room; a presence map generation unit that generates a presence map indicating the distribution of detection points corresponding to persons detected in images captured at different times; and an inferring unit that infers information regarding the room based on the presence map. The person detection unit can detect the face, head, or upper body of the person in the image, and acquire the position of the person in the room based on the position and size of the face, head, or upper body in the image. The inferring unit can infer that a polygon circumscribed around the distribution of detection points in the presence map is the shape of the room.

Detection of a specific person with a specific person detecting system using surveillance cameras installed in various environments has not been successfully accurate. A monitoring system of the present invention includes an image pickup device capturing images, an updating type specified person detection server receiving captured images, extracting image feature quantities therefrom and storing image feature quantities and provisional person information of unspecified persons and the reliability thereof in real time, and a monitor station with an image searching function capable of searching an image afterward, which are connected via a network. The updating type specified person detection server adds the person information to a specified person list during system operation on the basis of an unspecified person list that is updated.

A projection type video display apparatus in which a display device transmits light emitted from a light source to project a video written into the display device is characterized by comprising person detection means for detecting a person who exists within a projected video area, and control means for changing the state of the projected video when the person detection means detects the person.

A Collaborative Deep Network model method for pedestrian detection includes constructing a new collaborative multi-model learning framework to complete a classification process during pedestrian detection; and using an artificial neuron network to integrate judgment results of sub-classifiers in a collaborative model, and training the network by means of the method for machine learning, so that information fed back by sub-classifiers can be more effectively synthesized. A re-sampling method based on a K-means clustering algorithm can enhance the classification effect of each classifier in the collaborative model, and thus improves the overall classification effect. By building a collaborative deep network model, different types of training data sets obtained using a clustering algorithm are used for training a plurality of deep network models in parallel, and then classification results, on deep network models, of an original data set are integrated and comprehensively analyzed, which achieves more accurate sample classification.

The invention discloses a rail transit abnormal person detection method based on action recognition, and belongs to the field of aviation monitoring. The method comprises: inspecting railways by usingan unmanned aerial vehicle, extracting frames of videos, training and using an SSD detection model, obtaining position information of each person in each video frame image, intercepting a local areacontaining the person, and a key point detection model and a Reset-are detected. The 18-clothes classification model is trained; predicting joint coordinates of each person by using the key point detection model, forming a human skeleton sequence by using the joint coordinates in a certain time period, inputting the human skeleton sequence into an LSTM action recognition model, and recognizing theaction category of each person; and performing clothing classification of the personnel through the Resnet-18 clothing classification model; and judging whether the person is a worker or not according to the action type of each person and the corresponding appearance clothes. The method solves the problem of high cost of traditional manual inspection or roadbed inspection, and has high practicability and robustness.

A video segmentation method for segmenting video clips according to content of the video clips is disclosed. The method comprises scanning pixels of video frames with a first horizontal scan line to determine if colors of the pixels fall within a predetermined color range; creating a color map utilizing pixels located on the first horizontal scan line from a plurality of successive video frames; labeling the current video segment as a candidate video segment if the color map indicates the presence of a stable region of pixels falling within the predetermined color range for a predetermined number of successive video frames; and performing histogram color comparisons on the stable regions for detecting shot transitions. Audio signals of the video clips may also be analyzed to further verify the candidate video segments.

The invention discloses a face recognition security system which comprises a face acquisition module, a face contrast server, an access control recognition control module, an early-warning display module and a stranger management system, wherein the stranger management system receives stranger data sent by the face contrast server, controls the storage quantity of strangers by configuring similarity range parameters and valid time, and performs the stranger security early-warning based on the threshold value of the stranger data record. Through the face recognition security system, under the situation that the social public security has an urgent demand on recognition of strangers, the technical problem of recognizing strangers from massive security pictures / video data can be solved, and the breakthrough of tens of millions of order of magnitude recognition second feedback technology can be realized, thus helping a security department make early warning in the field of public security, and nipping in the bud.

The invention provides a visual compiler-based person and pose detection method. The method mainly comprises the steps of performing data synthesis of scene description; learning a net through synthesized data; defining the net by adopting basic blocks; and performing joint locating by using a pose net. According to the process, the scene description is used as an input of a visual compiler, a pedestrian detection system and a pose estimation system are trained by calibrated real data annotations, and the net is learnt through the synthesized data; then the net is defined by using two basic units including a residual module and a spatial belief module; and finally pedestrians are located by using the pose net. According to the method, the detected annotations, body part positions and segmented mask codes can be automatically obtained, the pedestrians are located by using a video camera, and the pedestrians are subjected to pose estimation and activity analysis; and the influence of illumination, shielding and the like on the detection is reduced, so that the identification efficiency is effectively improved.

The invention discloses a video pedestrian detection method fusing multi-target tracking clues, and belongs to the field of image processing of computer vision.A pedestrian detection module (improvedFaster R-CNN) and a multi-target tracking module are included. According to the method, the feature pyramid is introduced into the feature extraction network, pedestrians are detected on more scales,features of different layers are fused to improve the prediction effect, and the detection rate and accuracy of small targets are improved. According to the invention, the multi-target tracking moduleis used for assisting pedestrian detection, the inter-frame information of the front and rear frames of the video is introduced, the detection stability can be improved, and the historical frame target track and the target number obtained through the tracking module are used for relieving the detection instability caused by deformation, shielding and targets located at the edge of the picture. According to the invention, the anchor point box is modified in the pedestrian detection network part, so that the pedestrian detection network part better conforms to the characteristics of pedestrians, thereby improving the pedestrian detection precision.

The invention relates to a multifunctional photoelectric detection instrument and a target position observing and determining method implemented by the same. The multifunctional photoelectric detection instrument comprises refrigerating infrared sensors, a laser range finder, a visible light channel, a GPS (global positioning system) module, a magnetic compass, a comprehensive information processing unit, a power module and a control panel, the comprehensive information processing unit is respectively connected with the refrigerating infrared sensors, the laser range finder, the GPS module, the magnetic compass, the display module and the control panel, and the visible light channel is connected onto the laser range finder. The method includes methods for observing a target via the visible light channel, a method for observing the target and determining target direction via the refrigerating infrared sensor and a method for automatically solving coordinates of the target position. The multifunctional photoelectric detection instrument and the target position observing and determining method have the advantages that the target is detected and positioned by a single person, and images, the direction, the distance and position information of the target are accurately provided in real time. Besides, the various sensors are integrated, a miniaturization design is adopted, portability is high, and single-person detection capacity and information acquisition capacity are improved.

A detection device has an animal detection part that detects an animal from a captured image, a person detection part that detects a person from the image, and a detection result output part. The detection result output part outputs a detection result including at least information indicating that a target body is detected when the animal detection part detects the animal from the image and when the person detection part detects the person from the image.

In masquerading determination processing, a masquerading determination unit reads image data representing an image of an identification target on which a striped pattern is projected from an image storage unit to extract the striped pattern appearing in a face region of the image represented by the read image data. Subsequently, the masquerading determination unit determines whether a stripe in the face region in the image is a straight line or not. When the stripe is a straight line, because the identification target is a plane object such as a photograph or an image display device so that it can be determined that the target is at least not a person himself, the masquerading determination unit determines that the target masquerades. On the other hand, unless the stripe is a straight line, because the identification target has a solid configuration having three-dimensional irregularities to have a possibility of being a person himself, the unit determines that the target might not masquerade.

The invention discloses an automatic automobile door and window locking system and an automatic automobile door and window locking method. The method comprises the following steps of 1, detecting the distance between an electronic key of an automobile and an automobile body by using a distance detection module, and when the distance is detected to exceed a preset range, outputting a triggering signal to a person detection module; 2, detecting whether a person is in the automobile or not by using the person detection module, and when no person is detected in the automobile, outputting a control signal to an automobile locking module; 3, after the automobile locking module receives the control signal of the person detection module, automatically locking automobile doors and automobile windows. According to the system and the method, the automobile doors can be automatically locked to avoid property loss of an automobile owner under the condition that a driver forgets to lock the doors.