803 results about "Posture recognition" patented technology

Systems, methods and computer readable media are disclosed for a gesture recognizer system architecture. A recognizer engine is provided, which receives user motion data and provides that data to a plurality of filters. A filter corresponds to a gesture, that may then be tuned by an application receiving information from the gesture recognizer so that the specific parameters of the gesture—such as an arm acceleration for a throwing gesture—may be set on a per-application level, or multiple times within a single application. Each filter may output to an application using it a confidence level that the corresponding gesture occurred, as well as further details about the user motion data.

A method and device is used to teach and simulate swing mechanics. A camera is coupled to a computer such that frames including an image of a golfer swinging a golf club may be captured. A key part of the image is identified by the computer in sequential frames captured from the camera, and the image is compared to swing mechanics of a known standard golfer. The position of the known standard golfer is capable of following the natural swing of a student golfer, and the method and device are capable of providing real-time feedback to the student golfer during the golfer's swing. An instant replay may be used to reinforce good swing mechanics. A marker attachment may be used to assist in tracking and analyzing the swing mechanics of the images. Parameters relating to club angle and apparent club length may be compared to a known standard for a plurality of reference swing positions in order to determine the swing mechanics of the student golfer during practice instruction and within a golf simulation game to determine ball trajectory. A ball launch monitor may be integrated with the image analysis, which improves the effectiveness of the method and device synergistically.

A gesture-based recognition system obtains a vehicle occupant's desired command inputs through recognition and interpretation of his gestures. An image of the vehicle's interior section is captured and the occupant's image is separated from the background, in the captured image. The separated image is analyzed and a gesture recognition processor interprets the occupant's gesture from the image. A command actuator renders the interpreted desired command to the occupant along with a confirmation message, before actuating the command. When the occupant confirms, the command actuator actuates the interpreted command. Further, an inference engine processor assesses the occupant's state of attentiveness and conveys signals to a drive assist system if the occupant in inattentive. The drive-assist system provides warning signals to the inattentive occupant if any potential threats are identified. Further, a driver recognition module readjusts a set of vehicle's personalization functions to pre-stored settings, on recognizing the driver.

The invention discloses a physical education assisting system and method based on human body posture recognition. The system comprises a depth shooting device, a standard action storage unit, a display terminal, an action comparison unit and a prompting device, wherein the depth shooting device is used for obtaining user action depth image sequences comprising movement action of a user following the demonstration action; the standard action storage unit is used for storing a standard action model and standard action demonstration image sequences corresponding to the standard action model; the display terminal is used for displaying the standard action demonstration image sequences for the user; the action comparison unit is used for comparing the user action depth image sequence at a scheduled time point or time period with the standard action model at the corresponding time point or the time period, and obtaining the difference between the user action and the standard action model; the prompting device is used for outputting action evaluation information or action correction prompting information to a display device according to the difference. According to the physical education assisting system and method, the user does not need to wear a special recognition device, the manufacturing cost of the system is lowered, and the user experience is improved.

Systems, articles, and methods perform gesture identification with limited computational resources. A wearable electromyography (“EMG”) device includes multiple EMG sensors, an on-board processor, and a non-transitory processor-readable memory that stores data and / or processor-executable instructions for performing gesture identification. The wearable EMG device detects and determines features of signals when a user performs a physical gesture, and processes the features by performing a decision tree analysis. The decision tree analysis invokes a decision tree stored in the memory, where storing and executing the decision tree may be managed by limited computational resources. The outcome of the decision tree analysis is a probability vector that assigns a respective probability score to each gesture in a gesture library. The accuracy of the gesture identification may be enhanced by performing multiple iterations of the decision tree analysis across multiple time windows of the EMG signal data and combining the resulting probability vectors.

The invention discloses a method and a device for reconstructing a three-dimensional face based on a single face image. The method comprises the following steps of: performing posture recognition on a face image by utilizing face structure priori knowledge, and estimating rotating direction and angle of a face plane by combining face measurement and projective geometry so as to set a rotating angle of the three-dimensional face; estimating the depth of two-dimensional characteristic points on the face image by adopting an artificial neural network, and acquiring three-dimensional coordinates of the characteristic points; converting a common three-dimensional face model into a specific model by adopting a Dirichlet free deformation algorithm; and mapping the three-dimensional face model by adopting characteristic points extracted from the two-dimensional image. Therefore, the true three-dimensional model is constructed through the single face image, the three-dimensional modeling accuracy is effectively improved, the three-dimensional modeling time is shortened and the three-dimensional modeling cost is reduced.

Disclosed are methods and systems for monitoring and correcting sitting posture of a user and for discouraging sedentary behavior. The system comprises a processor, a portable seat cover with embedded pressure sensors and an angle sensor, a notification device, and a non-transitory storage medium for storing program code. The program code, when executed by the processor, causes the processor to initiate an idleness timer to count how long the user has been sitting, monitor the user's sitting posture by receiving real-time sensor measurements; identify a sitting posture by applying a posture identification rule, generate a posture correction notification through the notification device if the identified sitting posture is not ergonomically correct, determine whether a sedentary threshold has been reached and generate a stand notification if the sedentary threshold has been reached. Embodiments of the present invention help improve sitting posture and reduce sedentary behavior to minimize associated adverse health effects.

The invention discloses a human body posture recognition method based on self-adaptive extension Kalman filtering, belonging to the field of body-area networks. The method comprises a model design step and a parameter design step. In the model design step, the angular speed and accelerated speed of the motion of a human body and the peripheral magnetic field intensity are collected by virtue of an inertial sensor through the characteristic that the motion angle of limbs of the human body can be reflected by a quaternion, and posture resolving is carried out based on a self-adaptive extension Kalman filtering method so as to obtain a posture quaternion. In the parameter design step, by utilizing a theoretical analysis and experiment method, a process noise covariance matrix is determined, and the value of a noise covariance matrix, a state initial value and an initial value of a state covariance matrix are measured, so that the continuous iteration of the self-adaptive extension Kalman filtering method can be realized, and the motion posture of the human body is continuously recognized in real time. The human body posture recognition method can be used as a human body posture recognition method in the fields of physical training, medical care, game design and the like.

The invention provides an indoor inertial navigation algorithm based on posture recognition and a step length model and relates to the field of indoor inertial navigation and positioning. An inertial sensor is fixed on a foot of a pedestrian and data is transmitted to a smartphone by utilizing Bluetooth; the step frequency and the step number of the pedestrian are calculated by adopting a multi-condition and zero-speed detection method and a posture angle detection method according to posture characteristics of the foot when the pedestrian walks; an inertial navigation calculation result of each step of the pedestrian is counted and classified and the step length model is established; postures and the step frequency of the foot of the pedestrian are used as reference to calibrate the inertial navigation calculation result in real time; electronic compass data of a dynamic complement algorithm based on slide mean filter is used as a real-time course angle, and whether the pedestrian has an inertial and usual movement posture or not is judged according to changes of the step frequency and the course angle of the pedestrian; the step length model is matched by taking a judging result as the basis. The problems of an indoor positioning and inertial navigation technology that error accumulation is caused by double integration so that positioning is inaccurate and the like are solved, and the accuracy of indoor long-distance navigation of the pedestrian is guaranteed.

A vehicle operator condition monitoring system is provided. In one embodiment, the vehicle condition monitoring system includes a physical condition monitoring unit such as a camera, a data processing system, a navigation system, and a communication system. The physical condition monitoring unit is mounted in a vehicle to monitor at least one physical trait of a vehicle operator. The data processing system is coupled to the physical condition monitoring unit to compare data acquired by the physical condition monitoring unit with rules indicating types of action to be taken. The navigation system is functionally coupled to the data processing system allowing the data processing system to determine the location of the vehicle and the communications system is functionally coupled to the data processing system to allow the data processing system to contact an appropriate entity should the physical condition of the vehicle operator warrant such contact.

The invention provides a training method of a posture recognition model and an image recognition method and device. The training method of the posture recognition model comprises: inputting a sample image marked with human body key points into a feature map model included in the posture recognition model, and outputting a feature map corresponding to the sample image; inputting the feature map into a two-dimensional model included in the posture recognition model, and outputting two-dimensional key point parameters for representing a two-dimensional human body posture; inputting a target humanbody feature map cut from the feature map and the two-dimensional key point information into a three-dimensional model included in the posture recognition model, and outputting three-dimensional posture parameters for representing a three-dimensional human body posture; constructing a target loss function by combining the two-dimensional key point parameters and the three-dimensional attitude parameters; and based on the target loss function, updating model parameters of the attitude recognition model.

The invention discloses a human body action recognition method based on a TP-STG framework, which comprises the following steps: taking video information as input, adding priori knowledge into an SVMclassifier, and providing a posteriori discrimination criterion to remove a non-personnel target; segmenting a personnel target through a target positioning and detection algorithm, outputting the personnel target in a target frame and coordinate information mode, and providing input data for human body key point detection; utilizing an improved posture recognition algorithm to carry out body partpositioning and correlation degree analysis so as to extract all human body key point information and form a key point sequence; a space-time graph is constructed on a key point sequence through an action recognition algorithm, the space-time graph is applied to multi-layer space-time graph convolution operation, action classification is carried out through a Softmax classifier, and human body action recognition in a complex scene is achieved. According to the method, the actual scene of the ocean platform is combined for the first time, and the provided TP-STG framework tries to identify worker activities on the offshore drilling platform for the first time by using methods of target detection, posture identification and space-time diagram convolution.

The invention discloses a human body posture recognition method and device based on skeleton key points, a storage medium and a terminal. The method comprises: acquiring video image frames in a videoimage sequence; inputting the video image frame into a preset skeleton detection model for detection; when a human skeleton is detected, extracting a skeleton key point set of each target person in the detection area; and inputting the skeleton key point set into a pre-trained posture judgment model to generate a posture recognition result. Therefore, by adopting the embodiment of the invention, the accuracy of target human body posture recognition can be improved.

A posture recognition apparatus recognizes instructions signified by postures of persons present in the surroundings, from images obtained with an image capture device. The posture recognition apparatus includes an outline extraction device that extracts an outline of a body which is a candidate for a person from the images; a distance calculation device that calculates a distance to the body being the candidate, from distance information of each pixel within the outline in the image; a search device that searches for a candidate for a hand of a person based on the outline and the distance to the body represented by the outline; and a posture determination device that determines an instruction corresponding to the relative position of the candidate for a hand and the outline, and outputs this determination result as a posture determination result.

The application relates to a fitness monitoring device with an altimeter. The invention describes a biology metering monitoring device and includes various technologies which can be applied in the device. Besides, the technology for being employed by the altimeter in the biology metering monitoring device is provided. In some embodiments, the technologies are provided: re-calibrating the biology metering monitoring device based on position data; employing data of the altimeter as the aid of posture recognition; and / or employing the data of the altimeter to manage an airplane mode of the biology metering monitoring device.

An example input system calculates a specified position on a screen of a display device, the position being specified by an operating device. The input system includes an attitude calculation section, an identification section, and a specified position calculation section. The attitude calculation section calculates an attitude of the operating device. The identification section identifies one of a plurality of display devices toward which the operating device is directed, based on the attitude of the operating device. The specified position calculation section calculates a specified position in accordance with the attitude of the operating device as a position on a screen of the display device identified by the identification section.

The invention relates to a mattress fitting human body curves and a height adjusting method of the mattress, and belongs to the technical field of home life articles. According to the height adjusting method, the headrest area height of the mattress can be adjusted, that is, the height adjusting method comprises the following steps: firstly, a sleeping posture recognition model is established based on statistical pattern recognition; secondly, obtained human body pressure data are input in the sleeping posture recognition model, and sleeping postures with the highest similarity are taken as an output result; finally, a controller controls the working state of a height adjusting device according to the output result of the sleeping posture recognition model, so as to ensure that the headrest area of the mattress can be adjusted up and down. The mattress provided by the invention has the advantages that not only can the pillow height requirements of different people be met, but also the different pillow height requirements of people lying on the side and on the back can be met; through appropriate height adjustment for a spondylopathy patient, recovery of the vertebral curve of the spondylopathy patient can be facilitated; for people always lowering or raising heads to work, spondylopathy can be prevented.

The invention relates to a driving state monitoring and alarming system based on Kinect human body posture recognition. The system comprises a joint information acquisition unit, a driving posture processing unit and an alarming unit. The joint information acquisition unit is used for acquiring the body section image of a driver, obtaining the basic joint point positions of the human body according to an image recognition algorithm, binding joint point position data to a virtual skeleton model, establishing a space coordinate system and completing joint recognition. The driving posture processing unit is used for analyzing the virtual skeleton model of which joint recognition is completed, monitoring position information of all the joint points to obtain real-time movement information of the driver and comparing the position information of the joints with preset dangerous movement characteristic information, and transmitting a corresponding state abnormality signal to the alarming unit. The alarming unit is used for alarming after receiving the abnormality signal transmitted by the driving posture processing unit.

The invention provides a target object property recognition method and device and relates to the field of image recognition technology. The method is executed by the adoption of a recognition network.The recognition network comprises a feature extraction network, a pedestrian re-recognition network, a pedestrian posture recognition network and a pedestrian property recognition network, wherein the pedestrian re-recognition network, the pedestrian posture recognition network and the pedestrian property recognition network are connected with the feature extraction network, and parameters of therecognition network are obtained through joint training by the pedestrian re-recognition network, the pedestrian posture recognition network and the pedestrian property recognition network based on the mode of sharing the feature extraction network. The method comprises the steps that the feature extraction network extracts feature information of a target object; and the pedestrian property recognition network determines properties of the target object based on the feature information. Through the target object property recognition method and device, the pedestrian property recognition network can acquire richer and more comprehensive feature information by means of sharing the feature extraction network with the pedestrian re-recognition network and the pedestrian posture recognition network, and therefore the accuracy and robustness of property recognition are effectively improved.

The invention provides a pair of intelligent gloves, method and system for calorie consumption and hand posture recognition. The pair of intelligent gloves comprises a processor module, a memory module, a communication module, an induction module, an early warning prompt module, a display module, a power module, a switch module and a cloud server. The memory module, the communication module, the induction module, the early warning prompt module, the display module, the power module, the switch module and the cloud server are respectively connected with the processor module. The induction module comprises a pressure sensor and an IMU. The pressure sensor and the IMU are respectively used for being responsible for sensing hand stress strength and IMU data in users' fitness process, collecting source data according to user strength training and sending the source data to the processor module. The processor module is responsible for preprocessing of the source data. According to the pair of intelligent gloves, in the strength training process, the detection accuracy rate of detected motions is high, the rate of false alarm is low, and alarm signals can be sent when the hand posture is deviated.

A trailer target detection system includes a camera located on a vehicle and arranged to capture video images of a trailer towed by the vehicle. The system includes an image processor processing the video images to detect a gesture by a user indicative of position of target on the trailer and identifying location of the target based on the detected gesture. A vehicle trailer backup assist system includes a camera and an image processor for processing images to detect a gesture and determine a command based on the gesture. The processor controls backing of the trailer based on the command.

The invention discloses a shadow-boxing movement judging and guiding system based on kinect and a guiding method adopted by the shadow-boxing movement judging and guiding system. The system comprises a kinect depth sensor data acquisition module, a skeleton information filtration module, a feature information extraction module, a posture judgment module, a key posture correcting and guiding module and a consistency and standard judgment module for a group of movements. With the adoption of the system and the method, the judgment and guidance for the shadow-boxing movements are realized, and due to the design of the kinect depth sensor, the depth information is provided while the image information is provided, and therefore, the accuracy for the movement posture recognition is enhanced.

The invention relates to intelligent detection devices, and discloses an intelligent method and system for body building posture recognition, evaluation, early-warning and intensity estimation. The method comprises the steps that firstly, a thin film pressure sensor and an IMU on a pressure sensor on intelligent gloves are used for collecting original data generated in strength training; secondly, the original data is transmitted to a processor module for data preprocessing; thirdly, the processor module transmits pressure senor and IMU data obtained through preprocessing to an APP of a mobile terminal through a communication unit in a wireless communication mode. The intelligent method and system for body building posture recognition, evaluation, early-warning and intensity estimation have the advantages that the work of measuring calories and recognizing hand postures is carried out based on the existing sensor technology, an effective data processing method and an effective algorithm model are adopted, the method and the system can be widely applied to strength training, good and beneficial references are provided for body building fans, and recommendation for a dietary recipe and nutrition balancing in a later stage can be provided.

The invention provides an autonomous navigation method of robot based on video monitoring. The method comprises deployment of a navigation environment and installation of more than one monitoring cameras, and all permissible moving ranges of the robot are monitored and covered by a monitoring device. The method comprises following steps that S1, a host acquires a background image where there is no target object through the monitoring cameras and marks a static obstacle; S2, a posture recognition pattern which can be monitored by the monitoring cameras is drawn on the robot; S3, a target position and a target posture are input on the host, and the host plans a route after recognizing the pattern and the obstacle and carries out analysis and processing according to the posture; and S4, the host controls the robot to move according to the obstacle and the planned route. According to the invention, by use of the independent monitoring device, through the operation and analysis of the host, the precision and the robustness of the positioning are improved.

The invention discloses a target object posture recognition method, and the method comprises the steps: obtaining a current video frame; detecting a preset key point of the target object in the current video frame, and obtaining preset key point information of the target object in the current frame; according to the preset key point information, judging whether the position change of the preset key points of the current target object in the current frame and the previous f frames meets a first preset attitude condition, and / or judging whether the position relationship between the preset key points of the current target object in the current frame meets a second preset attitude condition; if the preset posture condition is met, the current target object posture is recognized as a preset posture, f is a preset natural number, and the preset posture condition is set according to the position characteristics between the key points of the to-be-recognized posture of the target object. According to the method of the invention, the method can accurately recognize the small change of the posture, is wide in application range, is low in requirements for an image in a video frame, is high inposture recognition accuracy, and is small in false detection and missing detection of posture recognition.

The invention discloses a human posture recognition method based on skeleton feature point extraction mainly to solve the problem on how to recognize human postures using a single camera. The method is implemented by the steps as follows: (1) inputting a video; (2) preprocessing the video; (3) extracting a coarse skeleton; (4) extracting skeleton feature points; (5) classifying the skeleton feature points; (6) performing gesture recognition; and (7) outputting a gesture recognition result. According to the invention, a coarse skeleton is acquired based on the shortest distance from the pixels of a profile graph to the edge, then, skeleton feature points are calculated and classified based on the distance from a chest node, and finally, a gesture is judged according to the relative position information of the skeleton feature points and the chest node. Human actions can be judged without a lot of calculation under the condition that there is only video steam input of a single camera. The human posture recognition method has the advantages of less calculation, low hardware requirement, high accuracy, and strong adaptability.

The invention relates to an interactive entertainment system based on the human gesture recognition and a realization method thereof. The system comprises a video capture module which is used for obtaining video stream images, a background differential module which is used for separating human region from the image, a human gesture recognition module which is used for drawing the posture parameters of every part of human body according to the human region, and an entertainment platform which is used for interactive entertainment according to the human region and the posture parameters. The system of the invention carries out the real time video capture by the human body posture in the scene and realizes the human body posture parameter extraction, thus completing the application of the follow-up interactive entertainment.

The embodiment of the invention provides a human body posture recognition method, device and equipment and a storage medium. The method comprises the steps of obtaining a to-be-identified image; wherein the to-be-recognized image comprises imaging information of a human body; performing human body detection on the to-be-recognized image to obtain at least one part of human body information; performing multi-scale recognition on the human body posture key points in the to-be-recognized image to obtain multi-scale key points; performing multi-scale fusion on the multi-scale key points to obtaintarget key points; distributing the target key points to each part of human body information to obtain human body posture information, and completing human body posture recognition; wherein the humanbody posture information is a set formed by the posture of each part of human body information. Through the embodiment of the invention, the accuracy of human body posture recognition can be improved.