53 results about "Mean motion" patented technology

An object in a video sequence is tracked by object masks generated for frames in the sequence. Macroblocks are motion compensated to predict the new object mask. Large differences between the next frame and the current frame detect suspect regions that may be obscured in the next frame. The motion vectors in the object are clustered using a K-means algorithm. The cluster centroid motion vectors are compared to an average motion vector of each suspect region. When the motion differences are small, the suspect region is considered part of the object and removed from the object mask as an occlusion. Large differences between the prior frame and the current frame detect suspected newly-uncovered regions. The average motion vector of each suspect region is compared to cluster centroid motion vectors. When the motion differences are small, the suspect region is added to the object mask as a disocclusion.

An object in a video sequence of frames is tracked by object masks generated for frames in the sequence. Macroblocks are motion compensated. Blocks matching entirely within a prior-frame object mask are used to generate an average object motion. When the average motion is below a motion threshold, frames are skipped at larger intervals, but more frequent frames are processed when high motion occurs. When the macroblock best matches a prior-frame block that has the object's boundary passing through the block, the macroblock is uncertain and is sub-divided into smaller sub-blocks that are again motion compensated. Sub-blocks matching blocks within the object mask in the base frame are added to the new object mask for the current frame while sub-blocks matching a block containing the object boundary are uncertain and can again be sub-divided to further refine the object boundary. Frame skipping and adaptive-size blocks on the object boundary reduce computational load.

The invention relates to a method for generating free gaits for a four-footed bionic robot. All of the free gaits of the four-footed robot are programmed according to moving periods, and each moving period comprises a four-footed supporting stage and a stepping stage; at the four-footed supporting stages, four feet of the robot are all positioned at supporting phases, the center of gravity of the robot moves in the advancing direction and the side directions of the advancing direction, in the advancing direction, the robot moves the largest advancing distance allowed by a current state, and in the side directions, the moving distance of the robot is determined taking account of the stability and the energy consumption of the robot at the same time; at the stepping stages, one foot of the robot is positioned at a swinging phase, and the other three feet of the robot are positioned at the supporting phases; when the swing foot of the robot for a certain moving period comes in contact with the ground, the movement of the robot for the next moving period is programmed. Through the adoption of the method, the adaptability of the four-footed robot to terrain is effectively improved, the stability of the robot can be improved, the continuity of the movement of the robot is ensured, the energy consumption is reduced, and the mean movement speed is accelerated.

The invention discloses a video region-of-interest extraction method based on visual perception characteristics and encoding information, and relates to the field of video encoding. The video region-of-interest extraction method comprises the following steps of (1) extracting luminance information of a current encoding macro-block from a primary video stream, (2) identifying a space domain visual characteristic saliency region through an inter-frame prediction mode type of the current encoding macro-block, (3) using a mean motion vector, in the horizontal direction, of a previous encoding macro-block and a mean motion vector, in the perpendicular direction, of the previous encoding macro-block as dual dynamic thresholds, identifying a time domain visual characteristic saliency region according to the result of comparison among a motion vector, in the horizontal direction, of the current encoding macro-block, a motion vector, in the perpendicular direction, of the current encoding macro-block and the dual dynamic thresholds, and (4) defining a video interest priority through combination of the identification result of the space domain visual characteristic saliency region and the identification result of the time domain visual characteristic saliency region, and achieving automatic extraction of a region of interest of a video. According to the video region-of-interest extraction method, the important encoding basis can be provided for the video encoding technology based on the ROI.

The invention discloses a movie action scene testing method based on analysis of a story line development model, which comprises the steps that: a video is carried out with the operation of pretreatment; the scene length of each scene is calculated; the average action intensity of the scenes is calculated; a movie editing factor is calculated by utilizing the scene length and the average action intensity of the scenes; the short-term audio energy of each audio frame and the average audio energy of the scenes are calculated; the average action dispersity of the scenes is calculated; a human perception factor is calculated by utilizing the average audio energy of the scenes and the average action dispersity of the scenes; according to the movie editing factor and the human perception factor, the story line development model is established and a story line development flow graph is generated according to the time order; action scenes in a movie are tested according to the story line development model. The testing method has the advantages that the story line development model is established by taking two viewpoints of the movie editing method and human perception into account to consider visual and auditory factors so as to simulate story line development and change, thereby realizing the accurate test of the action scenes in the movie.

The invention relates to a fighting detecting method based on stereoscopic vision motion field analysis. The method is as follows: an object human body is extracted according to the motion vector field and the depth map of the object; by taking the motion vector field as the main characteristic, the entropy of a cumulative histogram within a certain time is calculated out based on a block statistics cumulative motion vector direction histogram; a motion intensity degree evaluation strategy is planned according to the depth information of the object by calculating the average motion vector intensity in the block; and finally, the disorder and intensity degree evaluation scores are combined to obtain the probability of fighting behavior, through the time and space accumulation, when the probability of the fighting behavior is continuously greater than a threshold value, then the fighting behavior is confirmed. The fighting detecting method can filter the motion vector calculation errors caused by scene illumination variation and object blocking under the monitoring environment. The intensity degree judgment is more reasonable according to the scene depth information, and the robust detection result is obtained after the analysis of the time and space distribution by establishing the fighting probability model.

The invention discloses a water quality monitoring method based on motion features of an aquatic organism on a picture. A carassius auratus is taken as an indicator organism, two cameras are employed to shoot a swimming picture sequence of the carassius auratus in a fish tank and mean motional velocity, steering number of times, average angle of turning, average inclination angle of the body and time for staying at the upper part of water in a unit time are solved according to the picture. Water quality is automatically judged according to the unusual changes of the five indicators so as to alarm when wastewater is detected to have an exceeding concentration. The method comprises the following steps of: firstly, employing two cameras to shoot the swimming picture sequence of the carassius auratus in the fish tank, detecting target area information of the carassius auratus on the pictures by applying an image segmentation technology in two picture sequences, solving positions and angles of the carassius auratus on the frame pictures according to the target area information, then solving the mean motional velocity, the steering number of times, the average angle of turning, the average inclination angle of the body and the time for staying at the upper part of water in a unit time according to the positions and angles information of the carassius auratus on all the frame images and judging pollution degree of the water according to the five indicators finally. The water quality monitoring method based on the motion features of an aquatic organism on a picture has the advantages of simplicity and easy implementation and solves the problems of the unicity of physical and chemical testing in water quality monitoring and the like.

A method of and an apparatus for converting a frame rate using time shifting and motion compensation. The frame rate converting method includes calculating an average motion vector magnitude between every two adjacent frames among a predetermined number of frames input in series determining a frame section corresponding to a maximum value among the average motion vector magnitudes calculated every two adjacent frames and generating a new frame by performing motion compensation, to which a predetermined proportion is applied, in the frame section having the maximum motion vector.

A motion estimator and motion extracting method are provided for dividing a current frame into a plurality of blocks and estimating a motion comparing a current block with a previous frame. The motion estimator includes a mean motion estimation unit for computing a mean motion vector based on motion vectors of surrounding blocks adjacent to the current block, a line motion generation unit for generating a line motion vector in a predetermined search area based on motion vectors of horizontal blocks of the current block, and a motion vector selection unit for selecting and outputting one of the mean motion vector and the line motion vector as a final motion vector of the current block based on a mean motion expectation error value in accordance with the mean motion vector and a line motion expectation error value in accordance with the line motion vector.

The invention provides a multi-user precoding robust construction method under explicit delayed feedback, which is suitable for the multi-user parallel transmission based on the explicit feedback in a wireless multi-antenna transmission system. The method comprises the step that a base station researches all users serviced by the base station to obtain the mean motion speed of each user, computes the Doppler translation of each user according to the speed, computes the time-lapse related coefficient of each user according to a Clark scattering model, forms a delayed coefficient matrix, and substitutes the quantitative channel information fed back by a plurality of user sides and the obtained delayed coefficient matrix into a robust precoding construction formula provided by the method to obtain the precoding. The implementation process of the method in the practical application comprises the steps: 1) channel estimation; 2) delayed feedback; and 3) robustness precoding computation. The application efficiency can be higher if the precoding construction scheme provided by the method is adopted.

The invention relates to the field of the adjustment technology of a silicon wafer and provides a method and a device to use a photoelectric switch for dynamic adjustment of the deviation of the silicon wafer. The basic detection principle of the method is that: in the process of using a mechanical hand to deliver the silicon wafer to a reaction chamber from a transfer chamber, the transfer time of the silicon wafer is judged by a photoelectric switch sensor, and the operation distance is calculated by the combination of the average motion speed of the mechanical hand, so as to obtain the center coordinate of the silicon wafer on the current mechanical hand, the comparison of the value and the center coordinate of the silicon wafer without the deviation is carried out, and the obtained value difference is the deviation adjustment value which is needed. The invention has the advantage and the beneficial effect of accurately calculating the position of the deviation of the silicon wafer and guiding the mechanical hand to adjust the deviation of the silicon wafer automatically.

The invention relates to an intelligent analyzing method for passenger flow motion, which comprises the following steps of: (1) acquiring the average motion speed and the bidirectional passenger flow volume of pedestrians in real time through arranging a plurality of monitoring cameras; (2) carrying out intelligent analysis on pictures of all the monitoring cameras to generate a group of passenger flow statistical data; converting and analyzing all groups of data and comparing the data with a passenger flow motion rule in a danger sign state; and (3) sending a prewarning signal to a potential dangerous area. The intelligent analyzing method for passenger flow motion in the invention can be used for prewarning in time through acquiring the average motion speed and the bidirectional passenger flow volume of the pedestrians in real time and comparing the data with the passenger flow motion rule in the danger sign state, and informing a safety administration department to adopt measures in time so as to reduce risks causing accidents due to the overcrowding of people.

The embodiment of the invention provides a speed control method and device for a running machine and belongs to the technical field of running machines. The speed control method comprises the steps that the three-dimensional spatial data of two knee-joints of a runner within the preset time are obtained; the current motion amplitude parameter value of the runner is obtained on the basis of the three-dimensional spatial data of the two knee-joints of the runner within the preset time and compared with the average motion amplitude parameter value to obtain the acceleration or deceleration will of the runner; and the speed of the running machine is controlled according to the acceleration or deceleration will of the runner. In this way, the runner can conveniently, quickly and freely control the speed of the running machine according to the acceleration or deceleration will, so that the convenience and adaptability of the running machine are improved.

A method of correcting errors in motion vectors due to camera panning includes identifying an error block that contains an error motion vector out of a plurality of blocks in a current video frame, each block in the current video frame containing a corresponding motion vector. The method also includes calculating an average motion vector value of the motion vectors for all blocks in the current video frame that are not error blocks, identifying a group of effective motion vectors in the current video frame having corresponding motion vector values within a threshold value of the average motion vector value, calculating a panning motion vector value to be equal to an average value of the effective motion vectors, and correcting the error block by setting the motion vector of the error block to be equal to the panning motion vector value.

An object in a video sequence is tracked by object masks generated for frames in the sequence. Macroblocks are motion compensated to predict the new object mask. Large differences between the next frame and the current frame detect suspect regions that may be obscured in the next frame. The motion vectors in the object are clustered using a K-means algorithm. The cluster centroid motion vectors are compared to an average motion vector of each suspect region. When the motion differences are small, the suspect region is considered part of the object and removed from the object mask as an occlusion. Large differences between the prior frame and the current frame detect suspected newly-uncovered regions. The average motion vector of each suspect region is compared to cluster centroid motion vectors. When the motion differences are small, the suspect region is added to the object mask as a disocclusion.

The application discloses a vehicle-mounted video image stabilization method. The method comprises: extracting feature points in a video frame sequence; matching each two adjacent frames of feature points in the video frame sequence to obtain a set of matched feature point pairs; calculating a matrix parameter of a camera motion model; calculating a horizontal average motion vector of horizontal displacement A and a vertical average motion vector of vertical displacement; calculating a to-be-filtered horizontal motion vector according to the horizontal average motion vector and calculating a to-be-filtered vertical motion vector according to the vertical average motion vector; and carrying out motion compensation on the video frame sequence by using the a to-be-filtered horizontal motion vector and the to-be-filtered vertical motion vector. Therefore, random jittering between successive frames in a video frame sequence is compensated; horizontal and vertical movements occurring during a video shooting process are suppressed; the moving track of the video frame sequence after image stabilization becomes smooth; the vehicle-mounted video quality is improved; and the post processing of the vehicle-mounted video is realized smoothly.

The invention provides a bionic vision platform driven by a multi-axis linear motor, including a host, a motion control card, a motion mechanism, a camera and an image acquisition card. The above-mentioned parts are connected in series to form a closed loop, and the host controls the motion mechanism through the motion control card. , the motion mechanism drives the camera to rotate, the camera collects images, and outputs them to the host computer through the image acquisition card. The motion control card controls the movement of the vertical and horizontal linear motors. The mover of the horizontal linear motor is vertically fixed to the stator of the vertical linear motor. The connecting block is connected to the mover of the vertical linear motor. Two connecting rods are arranged in parallel. One end of the connecting rod is connected with the connection block, and the other end is connected with the camera. The platform controls the two degrees of freedom of the bionic eye through two linear motors, and the rapid movement of the two linear motors ensures that the camera can capture dynamic objects and targets, and can well complete various posture adjustments of the bionic eye, and is capable of dealing with complex work requirements in the environment.

The invention provides a method for measuring movement velocity of a high-density plasma and an optical system. On the basis of a principle of light refraction occurrence at a high-density-gradient plasma boundary, an optical system is designed to generate two light beams with incidence at small angles, wherein the directions of the two light beams are perpendicular to the movement direction of the plasma boundary; and a 4f system is used for imaging and separation of the two light beams: a first light beam and a delay beam is realized at a focal point of a first lens. When velocity measurement is carried out, some light of the first light beam is deflected at the boundary of the plasma in high-speed movement and the deflected light enters the optical system where the delay beam is located; an image of the boundary of the plasma at a corresponding time is displayed at a camera; after time delay of the delay beam, the shape and location of the plasma at the corresponding time are displayed at the camera; and average movement velocity information of the plasma boundary at a corresponding time interval can be obtained in a shadow image. With a laser shadow system, the velocity of the boundary of the plasma in high-speed movement can be obtained effectively; and the result is reasonable and accurate.

A safety system and method for a manual transmission vehicle with a remote starter is provided. The system includes a motion transducer module detecting motion of the vehicle and a controller module deciding erroneous starting of the vehicle and stopping the vehicle when erroneous starting is decided. The controller module receives motion data from the motion transducer module. A baseline is set in the detected motion data, and the controller module calculates number of baseline crossings that occur within a predetermined time frame in the motion data. The controller module decides erroneous starting based on the number of baseline crossings. The controller module adjusts the baseline based on averaged motion data from the motion transducer module when the remote starter is inactive.

The invention relates to a target tracking method of a mobile robot based on an electrical-hydraulic focusable lens. Modeling is carried out on an optical imaging system, a function relationship between a focal length of the electrical-hydraulic focusable lens and an optimal imaging object distance is obtained, distance information between a target in a clear state and the mobile robot and an average movement velocity of the target in a time interval t can be obtained by utilizing the function relationship, and by means of the information, a tracking speed of the mobile robot is updated, so that a distance between the mobile robot and the target is always kept within a proper range and effective tracking on the target is implemented; and meanwhile, due to automatic focusing on the target when the target distance information is acquired, an observer can always obtain an image feedback of the target, which is clear and rich in detail.

A safety system and method for a manual transmission vehicle with a remote starter is provided. The system includes a motion transducer module detecting motion of the vehicle and a controller module deciding erroneous starting of the vehicle and stopping the vehicle when erroneous starting is decided. The controller module receives motion data from the motion transducer module. A baseline is set in the detected motion data, and the controller module calculates number of baseline crossings that occur within a predetermined time frame in the motion data. The controller module decides erroneous starting based on the number of baseline crossings. The controller module adjusts the baseline based on averaged motion data from the motion transducer module when the remote starter is inactive.

The invention relates to a joint movement intelligent scoring method, comprising the following steps: collecting three-dimensional six-degree-of-freedom movement data of a joint of a tested person; calculating similarity scores of movement of each degree of freedom in the six-degree-of-freedom movement data; utilizing a DTW algorithm to compare the similarity of movement data of each degree of freedom of the joint of the tested person with average movement data of the corresponding degree of freedom of similar normal people, so as to obtain accumulated distance values of degree-of-freedom movement data of the joint of the tested person; mapping the accumulated distance values to calculate similarity scores of the degree-of-freedom movement of the joint of the tested person; and according to the similarity score of movement of each degree of freedom of the joint of the tested person, calculating the total score of six-degree-of-freedom movement of the joint of the tested person. Three-dimensional six-degree-of-freedom movement data of the joint are converted into the corresponding similarity scores, movement of each degree of freedom of the joint can be evaluated objectively, visually and accurately based on the scores, and comprehensive evaluation of three-dimensional movement of the joint can be realized through learning of the score of each degree of freedom.

The invention discloses a self-supervised learning method and device based on motion sequence regression. The method comprises the following steps: S1, randomly intercepting video sample fragments with a fixed frame number; S2, calculating to obtain the average motion amount of the video sample segment; S3, randomly disordering the sequence of the inner frames of the video sample fragments to obtain disordered video sample fragments; S4, performing sequential judgment on the out-of-order video sample fragments; S5, synthesizing the average motion amount of the video sample fragments obtained in the step S2 and the sequential evaluation result obtained in the step S4 to generate a final sequential score for the out-of-order video sample fragments, and taking the final sequential score as aregression target value. According to the method and the system, the information of motion sequence consistency in the video is fully utilized, and the sequential score is automatically judged by randomly generating the samples and the intra-frame sequence of the samples and establishing a sequential judgment standard, so that the purpose of automatic labeling is achieved.

The invention discloses a low earth orbit satellite broadcast ephemeris parameter expression method. The method comprises the steps that 1, broadcast ephemeris parameter fitting is performed based ona precise orbit of a low earth orbit satellite, wherein the expression method compatible with a navigation satellite broadcast ephemeris is adopted for broadcast ephemeris parameters, and satellite orbit semi-major axis variability, average motion angular velocity variability, latitude argument one-third circumferential sine and a cosine harmonic correction coefficient are added on the basis of original parameters to absorb periodic perturbation of the low earth orbit satellite orbit; and 2, reserved positions in a navigation message format of a navigation satellite are adopted to express theorbit semi-major axis variability, the average motion angular velocity variability, the latitude argument one-third circumferential sine and the cosine harmonic correction coefficient. According to the method, an existing medium and high earth orbit navigation satellite ephemeris design and expression method is fully utilized, parameter compatibility and expression mode unification of a low earthorbit satellite ephemeris and a medium and high earth orbit satellite navigation ephemeris are realized, and the complexity of comprehensive usage of the medium and high earth orbit navigation satellite ephemeris and the low earth orbit satellite ephemeris by a user is lowered.

A motion vector is searched for each block of the current video frame by using a previous average motion vector within a screen as a reference motion vector, and whether or not interpolation is effective is determined by using the obtained motion vector. If the interpolation is determined to be effective, an interpolation frame is generated between the current video frame and a preceding video frame by using the obtained motion vector.

Three slices, i.e., Basal, Mid, and Apical slices, which correspond to clinically useful ASE segmentation are designated, and the positions of the three slices are tracked through at least one cardiac cycle by performing three-dimensional speckle tracking in the remaining time phases. Three C-mode projection images concerning the tracked positions are reconstructed. In addition, arbitrary myocardial wall motion parameters at the tracked positions are computed and displayed upon being superimposed on C-mode images or projected / displayed on a polar map. As a C-mode projection image method, one of the following techniques can be used detecting and projecting only movement components perpendicular to slices determined in an initial time phase, detecting and projecting average movement components of the respective walls, and tracking and projecting each myocardial position. The obtained C-mode images are simultaneously displayed together with markers indicating the positions of long-axis images and C-mode images.

The invention relates to an infrared moving target detection method based on an adaptive neighborhood decision technology, which comprises the following steps: filtering an image sequence by using a side inhibition template, and filtering out a gently changed background and an enhanced target region; carrying out feature extraction on the image, taking the image subjected to feature extraction asan external input stimulus, obtaining a segmented binary image, and extracting a candidate target; performing connection component marking on each frame of image in the segmented binary image sequence, calculating the centroids of all connection components in each frame of image, and storing the centroids; respectively subtracting the maximum horizontal coordinate of the mass center of each frameof image from the maximum horizontal coordinate of the mass center of the previous frame of image to obtain a motion speed matrix of each frame of target; taking the median of the motion speed matrixnas the average motion speed of the weak targets in the image sequence, and calculating the mass center of the real target; and after the mass center of the real target is determined, finding a connected region where the mass center is located in the mark image, and extracting the target.

The embodiments of the invention provide a video-based flame detection method and an apparatus wherein the method comprises the following steps: calculating the motion amplitude and motion direction of each pixel point in the first motion region of a target video frame; according to the calculated motion amplitudes, calculating the average motion amplitude of the pixel points in the first motion region; based on the average motion amplitude, the motion amplitude and the motion direction of each pixel point, determining the number of the effective pixel points in each direction section; based on the numbers of the effective pixel points in all direction sections, calculating the dispersion degree of the effective pixel points; and according to the dispersion degree, determining whether the first motion region belongs to a flame region or not and obtaining the detection result based on the motion attributes. The technical schemes of the embodiments are capable of increasing the accuracy of a flame detection result.

The invention discloses a self-adaptive frame skipping method and a device. The method comprises the following steps: the mean value of the motion vector of a current frame is calculated; the weighted sum is used as a smoothing factor, and the weighted sum is composed of the ratio of the mean value of the motion vector of the current frame and the mean value of the average motion vector of at least one adjacent latest frame which is calculated as well as the ratio of the relative mean value of the motion vector of the current frame and the relative mean value of the average motion vector of at least one adjacent latest frame which is calculated; and if the value of the smoothing factor is within the range of the threshold value, the current frame is processed through frame skipping. The self-adaptive frame skipping device is also provided. The self-adaptive frame skipping method and the device accord with the HVS visual characteristics, ensure that a terminal user obtains optimal visual quality, and satisfy the image quality without affecting the visual quality when a user views the image.