MOTION CAPTURE FROM A MOBILE SELF-TRACKING DEVICE

    公开(公告)号:EP3400577A2

    公开(公告)日:2018-11-14

    申请号:EP16828825.6

    申请日:2016-12-15

    IPC分类号: G06T7/246 G06T19/00

    摘要: Methods and apparatus for capturing motion from a self-tracking device are disclosed. In embodiments, a device self-tracks motion of the device relative to a first reference frame while recording motion of a subject relative to a second reference frame, the second reference frame being a reference frame relative to the device. In the embodiments, the subject may be a real object or, alternately, the subject may be a virtual subject and a motion of the virtual object may be recorded relative to the second reference frame by associating a position offset relative to the device with the position of the virtual object in the recorded motion. The motion of the subject relative to the first reference frame may be determined from the tracked motion of the device relative to the first frame and the recorded motion of the subject relative to the second reference frame.

    VIDEO FRAME PROCESSING FOR MOTION COMPARISON

    公开(公告)号:EP3376436A1

    公开(公告)日:2018-09-19

    申请号:EP18164970.8

    申请日:2017-07-13

    IPC分类号: G06K9/00 G06T7/269 G06K9/62

    摘要: A method is disclosed of processing a sequence of video frames showing motion of a subject to compare the motion of the subject with a reference motion. The method comprises storing at least one reference motion data frame defining a reference motion, each reference motion data frame corresponding to respective first and second reference video frames in a sequence of video frames showing the reference motion and comprising a plurality of optical flow vectors, each optical flow vector corresponding to a respective area segment defined in the first reference video frame and a corresponding area segment defined in the second reference video frame and defining optical flow between the area segment defined in the first reference video frame and the area segment defined in the second reference video frame. The method further comprises receiving a sequence of video frames to be processed. The method further comprises processing at least one pair of the received video frames to generate a motion data frame defining motion of a subject between the pair of received video frames. Each pair of received video frames that is processed is processed by, for each area segment of the reference video frames, determining a corresponding area segment in a first video frame of the pair and a corresponding area segment in a second video frame of the pair. Each of the pairs of received video frames is further processed by, for each determined pair of corresponding area segments, comparing the area segments and generating an optical flow vector defining optical flow between the area segments. Each of the pairs of received video frames is further processed by generating a motion data frame for the pair of received video frames, the motion data frame comprising the optical flow vectors generated for the determined pairs of corresponding area segments. The method further comprises comparing the at least one reference motion data frame defining the reference motion to the at least one generated motion data frames defining the motion of the subject and generating a similarity metric for the motion of the subject and the reference motion.

    GESTURE RECOGNITION METHOD AND DEVICE
    7.
    发明公开

    公开(公告)号:EP3358495A1

    公开(公告)日:2018-08-08

    申请号:EP16856896.2

    申请日:2016-10-19

    IPC分类号: G06K9/00 G06K9/62 G06F3/01

    摘要: A gesture recognition method and an apparatus are provided. The method is applied to a terminal device based on a mobile communications network signal, the terminal device stores training sample data, and the training sample data includes a correlation spectrum characteristic vector of a reflected signal of a sounding reference signal SRS and a corresponding user gesture identifier. The method includes: transmitting, by the terminal device, a first sounding reference signal SRS; receiving, by the terminal device, a multipath reflected signal that is of the first sounding reference signal SRS and that is reflected from a gesture-making object, and performing a related operation on the reflected signal and a signal template corresponding to the first sounding reference signal SRS, to obtain a correlation spectrum characteristic vector of the reflected signal; and matching, by the terminal device, the correlation spectrum characteristic vector of the reflected signal against the training sample data, to recognize an input gesture.