For each frame of a video, a determination is made whether an image of a hand exists in the frame. When at least one frame of the video includes the image of the hand, locations of the hand in the frames of the video are tracked to obtain a tracking result. A verification is performed to determine whether the tracking result is valid in a current frame of the frames of the video. When the tracking result is valid in the current frame of the video, a location of the hand is tracked in a next frame. When the tracking result is not valid in the current frame, localized hand image detection is performed on the current frame.

A graphics processing system includes a processing circuit operable to render or decode a sequence of frames and generate extrapolated frames by extrapolating object motion from rendered or decoded frames. The system also includes a processing circuit operable to extrapolate object motion from first and second rendered or decoded frames in the sequence to a later extrapolated frame. The processing circuit is also operable to test candidate motion vectors from a region of the extrapolated frame through a region of the first frame to a region of the second frame by comparing the region of the first frame with the region of the second frame. A similarity measure from the comparison is used to select a motion vector and an indication representative of the selected motion vector is stored.

A graphics processing system includes a processing circuit operable to render or decode a sequence of frames and generate extrapolated frames by extrapolating object motion from rendered or decoded frames. The system also includes a processing circuit operable to extrapolate object motion from first and second rendered or decoded frames in the sequence to a later extrapolated frame. The processing circuit is also operable to test candidate motion vectors from a region of the extrapolated frame through a region of the first frame to a region of the second frame by comparing the region of the first frame with the region of the second frame. A similarity measure from the comparison is used to select a motion vector and an indication representative of the selected motion vector is stored.

A single-chip solution with smart recognition by feature extraction of embedded image sensor, which provides an image sensor that performs image recognition. The chip combines an image sensor, an image preprocessor, a motion detector, a scaler array, a feature extractor, a microcontroller, and memory onto a single silicon die. Only required external non-passive component is a serial flash storage device. The motion module detects object in motion. The data is fed back to sensor for optimal exposure on object in motion. The array of image scalers crops areas of interest from input. Cropped images are extracted by the HOG feature extractor into feature vectors. Feature vectors are compared against pre-trained for image recognition, or used to train the system to recognize new images. And the microcontroller can take the classifier result and perform application-specific tasks with it to become a smart application providing excellent progressive and economic value.

A single-chip solution with smart recognition by feature extraction of embedded image sensor, which provides an image sensor that performs image recognition. The chip combines an image sensor, an image preprocessor, a motion detector, a scaler array, a feature extractor, a microcontroller, and memory onto a single silicon die. Only required external non-passive component is a serial flash storage device. The motion module detects object in motion. The data is fed back to sensor for optimal exposure on object in motion. The array of image scalers crops areas of interest from input. Cropped images are extracted by the HOG feature extractor into feature vectors. Feature vectors are compared against pre-trained for image recognition, or used to train the system to recognize new images. And the microcontroller can take the classifier result and perform application-specific tasks with it to become a smart application providing excellent progressive and economic value.

A person tracking method, comprising: acquiring N frames in units of time windows; acquiring, in time windows, tracking paths of a target person according to the N frames; and constructing continuous tracking paths by means of continuous time windows, so as to obtain the tracking results of the target person.

A person tracking method, comprising: acquiring N frames in units of time windows; acquiring, in time windows, tracking paths of a target person according to the N frames; and constructing continuous tracking paths by means of continuous time windows, so as to obtain the tracking results of the target person.

A graphics processing system includes a processing circuit operable to render or decode a sequence of frames and generate extrapolated frames by extrapolating object motion from rendered or decoded frames. The system also includes a processing circuit operable to extrapolate object motion from first and second rendered or decoded frames in the sequence to a later extrapolated frame. The processing circuit is also operable to test candidate motion vectors from a region of the extrapolated frame through a region of the first frame to a region of the second frame by comparing the region of the first frame with the region of the second frame. A similarity measure from the comparison is used to select a motion vector and an indication representative of the selected motion vector is stored.

A graphics processing system includes a processing circuit operable to render or decode a sequence of frames and generate extrapolated frames by extrapolating object motion from rendered or decoded frames. The system also includes a processing circuit operable to extrapolate object motion from first and second rendered or decoded frames in the sequence to a later extrapolated frame. The processing circuit is also operable to test candidate motion vectors from a region of the extrapolated frame through a region of the first frame to a region of the second frame by comparing the region of the first frame with the region of the second frame. A similarity measure from the comparison is used to select a motion vector and an indication representative of the selected motion vector is stored.

A motor vehicle vision system (10) includes a pair of imaging devices (12a, 12b) forming a stereo imaging apparatus (11) and a data processing apparatus (14) for rectification of images captured by the stereo imaging apparatus (11), matching of rectified images, and to detect an object in the surrounding of the motor vehicle. The data processing device (14) performs, for image elements (43) of a rectified image from one imaging device, a search for a best-matching image element (44) in the corresponding rectified image from the other imaging device. The search yielding vertical shift information from which a vertical shift from the image element (43) to the best-matching image element (44) is derivable. The data processing device (14) calculates a pitch angle error and/or a roll angle error of or between the imaging devices (12a, 12b) from the vertical shift information.