579results about How to "Stable tracking" patented technology

A high-speed marker-free motion capture, which is capable of powerfully detecting a body's feature points corresponding to a body's end portions such as a head, hands, feet, trunk, arms and legs at a high speed in an illumination change or background or noises of cameras. The extracted feature points of the body can be directly tracked stably in a 3-dimensional space. The position errors of the feature points due to the change of the illumination conditions or a shadow can be automatically corrected and the feature points can be stably tracked with respect to overlapping and disappearance of the feature points. Further, when coordinates of the middle joints are estimated using 3-dimensional coordinates of the extracted feature points of the body, the present invention restores a human model by estimating the positions of the middle joints of the actor with high accuracy without using a motion database, thereby securing the stability and reality of the 3-dimensional motion data required in the motion capture.

The invention discloses a transitional tracking planning method applied by an industrial robot, which can realize the transition between a joint space track and a cartesian space track, and the transition between two tracks of the cartesian space, the transitional track between different movement tracks are planned under the cartesian space, and has an intuitive shape; by adopting the algorithm that two parabolas are fused into one transitional curve, the smoothness of the track, speed and acceleration can be ensured, and the curve shape is controllable; the transitional track is formed by six independent curves, and the transition can be realized on the track with only posture change without position change; from the engineering application angle, the path velocity of transitional track boundary is restrained by utilizing the included angle between the tracks and the system allowable chord error, the boundary posture rotation speed is restrained by the similar mode, so that the large impact on a mechanical system caused by overhigh engagement speed can be prevented.

The invention discloses a multi-target tracking method based on semantic information and scene information, and the method comprises the steps: predicting the position of each tracking target in a scene in a current frame, carrying out the correction of the predicted position through employing a detection model, and obtaining a semantic score of the predicted position; taking the corrected position as a target frame, obtaining the similarity between the target frame and the historical track of the corresponding tracking target, and fusing the semantic score and the similarity to obtain the tracking score of the target frame; updating a scene model of the scene grid according to the tracking score of the target frame, calculating a scene confidence coefficient of the target frame accordingto the scene model, and updating the tracking score of the target frame according to the scene confidence coefficient; and obtaining a detection result of the current frame by using the detection model, matching the target frame with the detection result, and determining the state of the tracking target or generating a new target according to the matching result and the tracking score of the target frame, thereby obtaining a tracking result of the current frame. According to the invention, the robustness and accuracy of multi-target tracking can be improved.

The invention discloses a video monitoring system and a video monitoring method. Direction of a pan-and-tilt camera is adjusted according to received pan-and-tilt control commands, and the pan-and-tilt camera acquires front-end videos and outputs the same to a pan-and-tilt camera path. A monitoring platform feeds the pan-and-tilt camera path back to a client side and tracks a policy list automatically after performing access authentication on the client side, and outputs received pan-and-tilt control commands to the pan-and-tilt camera after performing control authentication on the received pan-and-tilt control commands. The client side selects an automatic tracking policy from the automatic tracking policy list, packages presetting-bit commands in the pan-and-tilt control commands and outputs the pan-and-tilt control commands to the pan-and-tilt camera; and according to the acquired pan-and-tilt camera path, real-time videos are downloaded to play, and according to the played videos, a target object is positioned, and the position of the target object at the next calculating time is predicted and packaged in the pan-and-tilt control commands to transmit. By the video monitoring method, tracking efficiency can be improved while monitoring cost is reduced.

The invention discloses a method and a system for automatically tracking a moving pedestrian video based on particle filtering. The method comprises the following steps of: inputting one frame of images, and carrying out detection through an HOG (Histogram of Oriented Gradient) feature vector set and an SVM (Support Vector Machine) vector machine; in order to realize particle filtering tracking based on double HOG and color features, firstly obtaining an initial rectangular area of target pedestrian, sampling a plurality of particles from a target rectangular area, extracting an HOG feature and a color feature, computing the weight of the particles after the double HOG and color features are fused, obtaining the final state estimation through a minimum mean square error estimator, outputting an estimation target and then resampling; and closely locking the tracked target pedestrian. The method extracts the double HOG and color features to increase the robustness of a particle filtering likelihood model and eliminate the unstable situation in the tracking process, the method combines the HOG feature to build the better likelihood model through a fusion strategy of weighted mean, the robustness of the tracking algorithm is greatly increased, and the stable tracking is completed.

A radar device tracks with a high accuracy positions and velocities of a plurality of external targets that are close to each other and whose observed direction values are likely to be low. The radar device includes: a target tracking filter for calculating relative distances and relative velocities of a plurality of external targets by signal-processing received signals from an antenna, for calculating the directions of the plurality of external targets by combining, among beam patterns radiated by the antenna, adjacent beam patterns that partially overlap, and for obtaining, from the directions and the relative distances and velocities, observed position values and observed velocity values of the plurality of external targets, to calculate, from the observed position values and the observed velocity values, smoothed values of the position and velocity for each of the external targets; and an intra-tracking-processing-cluster target tracking filter for forming a cluster from the plurality of external targets that are close to each other, for creating gates for the external targets in the cluster, different from those in the target tracking filter, and for performing a correlation process on the observed values of the external targets based on the gates.

The invention discloses a continuous trajectory planning transition method for a robot tail end. The method comprises following steps that firstly, a first line segment and a second line segment which need to be subjected to continuous trajectory planning transition are determined, demonstration points of a connecting line segment are determined, and the transition distance between the demonstration points and the first line segment and the transition distance between the demonstration points and the second line segment are determined; secondly, according to the transition distance between the demonstration points and the first line segment, first transition joint points on the first line segment are determined, and according to the transition distance between the demonstration points and the second line segment, second transition joint points on the second line segment are determined; and thirdly, an amplitude coefficient, a phase coefficient and a speed zooming coefficient are calculated in a coordinate axis, and the amplitude coefficient, the phase coefficient and the speed zooming coefficient are brought into a limited term sine position planning function to determine a transition curve expression between the first transition joint points and the second transition joint points. The algorithm flow is clear, the calculating time is greatly shortened, and the complex degree of a robot control system is reduced.

A method for monitoring and securing a facility without employing a manned control center, according to which a server acquires surveillance data from each of a plurality of distributed surveillance devices deployed within or adjacent to a given security sensitive facility. Then the server receives an intrusion alert from a guard, or from one or more of the plurality of surveillance devices to initiate a triggering event, if the acquired surveillance data is indicative of activity that has been detected at an indicated location. A first guard is alerted and presented with a full situational awareness image to assess the severity of the detected activity and the triggering event is terminated if found by the first guard to be of a low severity, or an interception operation initiated by self-dispatching or dispatching one or more additional guards to a close by location that holds tactical advantage relative to the indicated location if the triggering event was found by the first guard to be of a high severity.

The invention discloses a multi-feature and Kalman filtering combined object tracking method for realizing the effect of object tracking under a complicated background environment and a sheltering condition. Comprehensive feature based on color, texture and object moving information is provided for improving a CamShift algorithm, and a Kalman filter is combined to predict an object moving state, so that the tracking stability and the tracking accuracy of the moving target under a complicated background are improved. Color information is most likely to be interfered by factors such as illumination and background interference color; object texture feature is introduced to improve the defect when single color information is adopted; target moving information is continuously introduced; and further the interference in the background is removed. When an object is sheltered, least square fitting and target moving track extrapolation are performed through priori information before object shelter, the object moving positional information is predicted, and recapture of a moving object when the shelter is finished is facilitated.

The invention discloses a correlation filtering and deep twin network-based robust long-range target tracking method, and relates to a computer vision technology. According to the method, correlationfiltering and a deep twin network are combined under a uniform tracking framework, so that challenges of target occlusion and view disappear in long videos can be effectively processed. In the tracking method, a D-expert and C-expert-based expert assessment mechanism is capable of effectively assessing and screening target candidate positions jointly generated by the correlation filtering and thedeep twin network so as to obtain optimum target tracking results, and the results are used for updating correlation filtering trackers, so that the condition that the correlation filtering trackers are updated by wrong samples is effectively avoided. The target tracking method is relatively robust for various challenges in long videos, and is capable of stably tracking targets for a long time.

In each of track links (14) forming a crawler belt (13), an engaging projection (23) engaging to a drive wheel (11) is provided on a drive wheel abutting surface (22) abutting on the drive wheel (11). A fitting groove (27) composed of a bottomed recessed groove is provided on a ground contact surface (26) at the opposite side to the drive wheel abutting surface (22) in the upper-lower direction in a position corresponding to the engaging projection (23). Thereby, at the time of stacking up the track links (14) in the upper-lower direction, the fitting groove (27) of the upper track link (14) can be fitted into the engaging projection (23) of the lower track link (14).

The invention discloses a multi-feature fusion multi-target tracking method based on Kalman filtering assistance, and the method comprises the steps: firstly, reading any two frames of images in a video frame, inputting a preprocessed image into a multi-target detector, and obtaining a detection result of each frame in a video; introducing a target occlusion mechanism, enabling the judgment mechanism to carry out judgment according to coordinates of a target center point and the size of a target, and if the occluded part is small or no occlusion exists, enabling the detector to input the masscenter coordinates of the detection frame and the preprocessed video frame into a pre-trained convolutional neural network, extract the semantic information of the shallow layer and the deep layer ofthe target, perform cascading to form a feature matrix, and then perform similarity estimation on the feature matrixes of the two frames to obtain an optimal track; and if the detected target occlusion condition is serious, inputting the mass center coordinates of the detection frame into a Kalman filter, estimating the position information of the target in the next frame according to the previousmotion state of the target, and comparing the estimated coordinate information with an actual detection result to obtain an optimal track.

The invention discloses an adaptive anti-occlusion infrared target tracking method based on multi-layer depth feature fusion. Firstly, a series of multi-layer depth feature maps of the same size and different levels are obtained; the multi-layer depth feature map is then converted from a time domain to a frequency domain according to correlation filtering, a filter is trained and the response mapis calculated according to the fast Fourier transformation, and then the multi-layer depth feature map is merged and dimensionally reduced according to the weighted fusion of the intra-layer features,so that the feature response map of different levels is constructed and the maximum correlation response value is obtained, which is the estimated position of the target; at last, that dense featuresof the target are extracted, and the response confidence of the target center position estimated by the depth convolution feature is obtained according to the maximum response value of the feature obtained by the correlation filtering; when the response confidence of the target center position is less than the re-detection threshold T0, the obtained target estimated position is evaluated by on-line target re-detection and the position of the target is adaptively updated according to the evaluation result.

The envelope tracking power supply includes a plurality of variable voltage supplies. A power supply controller of a power amplifier controls the plurality of variable voltage supplies so as to precisely divide the power supply voltage in a high frequency area, based on the voltage distribution of a received transmitted base band signal, in order to maximize the power added efficiency of the power amplifier. The power supply controller includes a threshold memory including a plurality of first zones, and a frequency memory including a plurality of second zones. The power supply controller changes the threshold held in the first zones so that each second zone approaches the average of the second zones.

Controlling a power converter circuit for a direct current (DC) power source is disclosed. The power converter may be operative to convert input power received from the DC power source to an output power and to perform maximum power point tracking of the power source. The power converter is adapted to provide the output power to a load that also performs maximum power point tracking.

The embodiment of the invention provides a human flow statistics method, device and system. The method comprises a step of acquiring continuous frames of images collected by an image acquisition device, a step of inputting the continuous frames of images into a full convolutional neural network obtained by training and generating a head confidence distribution map of each frame of image in the continuous frames of images, a step of determining at least one human head detection target in each frame of image by using a preset target determination method for the head confidence distribution map of each frame of image, a step of obtaining a feature matching result and motion smoothness of any human head detection target in each frame of image, carrying out target association on the current frame and a previous frame of the current frame to obtain tracking targets according to the feature matching result and the motion smoothness of any human head detection target in each frame of image, distributing tracking identifiers to the tracking targets, and a step of counting the number of all tracking identifiers to obtain a result of human flow statistics. Through the method, the device and the system, the accuracy and operational efficiency of the human flow statistics can be improved.

The invention discloses a motion collecting and feedback method and system based on stereoscopic vision. The method comprises the following steps that the space position of an infrared light source in an attachment position in a front space is collected, the space position is mapped into a virtual space, and the position and direction of a virtual object are controlled; when the virtual object interacts with a virtual environment, a wireless control attachment device vibrates and simulates the feedback feeling of force. The motion collecting and feedback method and system based on stereoscopic vision have the advantages that positioning is rapid and accurate, and operation is stable; the hardware is quite low in cost, simple in structure and quite small in size; force feedback is achieved and can largely improve the sense of reality of interaction behaviors, and user experience is improved; the reality function is improved, and the man-machine interaction modes are increased.

The invention relates to a tracking method for the maximum power point, applied in an independent photovoltaic grid-connected power generation system. According to the tracking method, the maximum power point of a solar cell panel is tracked in a variable step size mode on the basis of the conductance increment method, so that the problem that speed and accuracy are ambivalent in a traditional conductance increment method is solved. According to the tracking method, the step size of a work point can be achieved in an automatic regulating mode according to characteristics of the solar cell panel, and therefore when the work point is far away from the maximum power point, the step size is larger, and tracking speed is increased; when the work point is close to the maximum power point, the step size is reduced so that the oscillation phenomenon nearby the maximum power point can be reduced, and therefore efficiency is improved. The step size correcting MPPT algorithm provided by the tracking method can solve the conflicting problem, and accuracy and speed of MPPT can be improved easily and effectively. The tracking method has the advantages of being accurate in control and high in response speed, and can be suitable for rapid reaction occasions where sunshine and temperature change suddenly, and the situation of continuous changes of the sunshine can be stably tracked.

Provided are a human-tracking method and a robot apparatus. The human-tracking method includes receiving an image frame including a color image and a depth image, determining whether user tracking was successful in a previous image frame, and determining a location of a user and a goal position to which a robot apparatus is to move based on the color image and the depth image in the image frame, when user tracking was successful in the previous frame. Accordingly, a current location of the user can be predicted from the depth image, user tracking can be quickly performed, and the user can be re-detected and tracked using user information acquired in user tracking when detection of the user fails due to obstacles or the like.

The invention belongs to the technical field of unmanned driving, and discloses a target following and dynamic obstacle avoidance control method for a speed difference slip steering vehicle, and the method comprises the steps: building four neural networks through employing a depth determinacy strategy in reinforcement learning; constructing a cost range of the obstacle so as to determine a single-step reward function of the action; determining continuous action output through an actor-critic strategy, and updating network parameters continuously through gradient transmission; and training a network model for following and obstacle avoidance according to the current state. According to the method, the intelligence of vehicle following and obstacle avoidance is improved, and the method canbetter adapt to an unknown environment and well cope with other emergencies. the complexity of establishing a simulation environment in the reinforcement learning training process is reduced. By utilizing a neural network prediction model trained in advance, the position and posture of each step of the target vehicle and the obstacle can be obtained according to the initial position and posture ofthe target and the obstacle and the action value of each step, so that the simulation accuracy and efficiency are improved.

The invention provides a solar concentrator and a condensing method. The solar concentrator includes: a concave cylindrical trough support surface and a condenser mirror array of unified structure matched on the cylindrical trough support surface. The condenser mirror array of unified structure includes a plurality of strip-shaped reflector mirrors; the length of each strip-shaped reflecting mirror is along the axial direction of the cylindrical trough support surface while the width is along the circumferential direction of the cylindrical trough support surface; and all strip-shaped reflecting mirrors are flexibly connected. The invention also includes an absorption device arranged above the cylindrical trough support surface. The condenser mirror array of unified structure is collocated into the shape of sunlight reflected into the size-restricted space of the absorption device; and the condenser mirror array of unified structure revolves the sun and rotates wholly along the circumferential direction of the cylindrical trough; in addition, in the rotating process, the condenser mirror array can maintain the shape of sunlight reflected into the size-restricted space of the absorption device.

The invention provides an object tracing method based on active scene learning, belonging to the technical field of computer graphics and image mode identification. The difficulty needing to be confronted all together is to not only guarantee the adaption ability of a tracing system to the target change, but also guarantee the learning accuracy to avoid shifting to result in tracing failure. According to the invention, the object tracing is achieved by the steps of online modeling, short-time tracing, overall detection, active scene learning, a constraint method, analysis and extraction of a motion area and the like. By learning a scene online, the background information can be found actively, meanwhile, the motion area analysis and extraction method based on light streams is provided, and the problem of target loss and difficulty in recovering the target, resulting from rapid movement of an objected or severe movement of the scene, can be effectively solved according to the structured constraint. The adaption ability to the target change can be effectively improved, so that the long-time stable rapid object tracing can be achieved, and the object tracing method is mainly applied to various object tracing occasions.

The invention discloses a moving target anti-fusion shielding tracking algorithm which specifically comprises the following steps of: firstly, judging whether the current moving target is in a fusion shielding state according to the target shielding judgment criterion; classifying all shielding targets according to the target moving information, and separating the active shielding targets and the passive shielding targets; obtaining the real positions of the targets by different template matching algorithms for different types of shielding targets respectively, wherein an overall template matching method based on fast Fourier transform and integral images is adopted for the active shielding targets, and an improved sub-block template matching method is adopted for the passive shielding targets; matching and estimating the moving displacement of the sub-block template by use of a target prospect adaptive block to realize the adaptive update of the template; and finally, realizing the anti-fusion shielding tracking of the target by combining the target overall template matching based on fast Fourier transform and integral images and the adaptive multi-sub-block template matching.