The method comprises the steps of, successively for each image following a preceding image based on a map of motion vectors that corresponds to the motion from said preceding image toward said following image, building a pixel mask for said following image, applying said pixel mask to the corresponding following image in order to obtain a corresponding masked image that samples the pixels of said following image. The application of this method to color clustering allows the iterative update of the colors clusters with limited computer resources.

The method comprises the steps of, successively for each image following a preceding image based on a map of motion vectors that corresponds to the motion from said preceding image toward said following image, building a pixel mask for said following image, applying said pixel mask to the corresponding following image in order to obtain a corresponding masked image that samples the pixels of said following image. The application of this method to color clustering allows the iterative update of the colors clusters with limited computer resources.

A method for pull frame interpolation includes receiving an encoded bitstream including information representing a plurality of frames of video data, decoding the plurality of frames, including identifying a plurality of motion vectors indicating motion from a first frame of the plurality of video frames to a second frame of the plurality of video frames, identifying an interpolation point between the first frame and the second frame, identifying a plurality of candidate interpolation motion vectors indicating motion from the first frame to the interpolation point and from the second frame to the interpolation point based on the plurality of motion vectors, selecting an interpolation motion vector from the plurality of candidate interpolation motion vectors based on a metric, and generating an interpolated frame at the interpolation point based on the selected interpolation motion vector, which may include correcting an artifact in the interpolated frame by blending the interpolated frame.

A method for pull frame interpolation includes receiving an encoded bitstream including information representing a plurality of frames of video data, decoding the plurality of frames, including identifying a plurality of motion vectors indicating motion from a first frame of the plurality of video frames to a second frame of the plurality of video frames, identifying an interpolation point between the first frame and the second frame, identifying a plurality of candidate interpolation motion vectors indicating motion from the first frame to the interpolation point and from the second frame to the interpolation point based on the plurality of motion vectors, selecting an interpolation motion vector from the plurality of candidate interpolation motion vectors based on a metric, and generating an interpolated frame at the interpolation point based on the selected interpolation motion vector, which may include correcting an artifact in the interpolated frame by blending the interpolated frame.

In a motion estimation for a group of at least one image element in a frame of a video sequence, a global motion is determined between the frame and a reference frame. Uncovered groups present in an uncovered region of the frame are identified based on the determined global motion. The global motion is assigned as motion representation for these identified uncovered groups. The assigned motion representation is useful for constructing new frames in the sequence in a frame rate up-conversion.

In a motion estimation for a group of at least one image element in a frame of a video sequence, a global motion is determined between the frame and a reference frame. Uncovered groups present in an uncovered region of the frame are identified based on the determined global motion. The global motion is assigned as motion representation for these identified uncovered groups. The assigned motion representation is useful for constructing new frames in the sequence in a frame rate up-conversion.

A proposed intermediate way of handling the renderable portion of the first view results in more efficient coding. Instead of omitting the coding of the renderable portion completely, even more efficient coding of multi-view signals entails merely suppressing the coding of the residual signal within the renderable portion, whereas the prediction parameter coding still takes place from the non-renderable portion of the multi-view signal across the renderable portion so that prediction parameters for the renderable portion may be exploited for predicting parameters for the non-renderable portion. The additional coding rate for transmitting the prediction parameters for the renderable portion may be kept low as this merely aims at forming a continuation of the parameter history across the renderable portion to serve as a basis for prediction parameters of other portions of the multi-view signal. Expressed differently, the prediction parameters for the renderable portion need not perfectly predict the texture within the renderable portion of the first view to keep the residual signal within the renderable portion low.

A proposed intermediate way of handling the renderable portion of the first view results in more efficient coding. Instead of omitting the coding of the renderable portion completely, even more efficient coding of multi-view signals entails merely suppressing the coding of the residual signal within the renderable portion, whereas the prediction parameter coding still takes place from the non-renderable portion of the multi-view signal across the renderable portion so that prediction parameters for the renderable portion may be exploited for predicting parameters for the non-renderable portion. The additional coding rate for transmitting the prediction parameters for the renderable portion may be kept low as this merely aims at forming a continuation of the parameter history across the renderable portion to serve as a basis for prediction parameters of other portions of the multi-view signal. Expressed differently, the prediction parameters for the renderable portion need not perfectly predict the texture within the renderable portion of the first view to keep the residual signal within the renderable portion low.

A method and apparatus for video encoding or decoding utilizing adaptive background residual prediction is disclosed. The present invention adaptively applies background residual prediction to a current block based on a selection decision. The coding block is split into one or more coding sub-blocks. A reference sub-block in a reference picture is located for a current coding sub-block of the current coding block according to a motion vector associated with the current coding block. A background reference sub-block in a background picture is located for the reference sub-block, where the background reference sub-block is at a first co-located location as the reference sub-block. The method then selects a first predictor or a second predictor to encode or decode the current sub-block based on a selection decision. The first predictor corresponds to the reference sub-block, and the second predictor is derived according to the reference sub-block and the background picture.

A method and apparatus for encoding a sequence of images by predicting at least one current image by forward motion compensation from at least one reference image. Prediction implements, for at least one current block of the current image: forward projection of a block of the reference image on the current image, delivering a projected block at least partially overlapping the current block; partitioning the current block delivering a sub-block at least partially overlapped by at least one projected block; and at least one iteration of the following steps for at least one overlapped sub-block: checking that the overlapped sub-block complies with a predetermined allotting criterion; with a positive check, allotting, to the overlapped sub-block, one of the projected motion vectors; and with a negative check, and so long as a predetermined minimum size is not reached for the overlapped sub-block: partitioning of the overlapped sub-block.