Video encoding and decoding with improved error resilience
For an image portion to encode an initial set of motion information predictors is obtained. It is tested whether the number of motion information predictors in the initial set is lower than a target number and, if so, one or more motion information predictors are added to the initial set to generate a set of motion information predictors having controlled diversity. The motion information predictors of the initial set are actual motion information predictors, having motion vectors obtained from image portions of the image being encoded or of a reference image, and potential motion information predictors for addition include one or more further such actual motion information predictors and also include one or more virtual motion information predictors not having motion vectors obtained from image portions of said image being encoded or of a reference image. A motion information predictor for the image portion to encode is selected from the generated set of motion information predictors.
1. A method of decoding a bitstream comprising an encoded sequence of digital images, at least one portion of an image being encoded with respect to a reference image by a motion estimation process that determines motion information identifying the reference image portion, the method comprising decoding the motion information using a motion information prediction process comprising:
obtaining, for a target image portion of an image to be decoded, if available, one or more actual motion information predictors having motion vectors obtained from motion vectors associated with first image portions of the image to be decoded wherein the first image portions are adjacent to the target image portion;
if a number of the one or more obtained motion information predictors is lower than a target number, wherein the target number is greater than one, adding one or more motion information predictors to generate a set of motion information predictors until a number of motion information predictors in the set is equal to the target number, wherein the adding of one or more motion information predictors comprises:
performing a first process of adding, if available, one or more further actual motion information predictors having motion vectors obtained from motion vectors associated with second image portions of the image to be decoded or of the different image from the image to be decoded; and
if a number of motion information predictors after the first process is lower than the target number, performing a second process of adding one or more virtual motion information predictors, the one or more virtual motion information predictors having motion vectors which are neither associated with image portions of the image to be decoded nor of the different image from the image to be decoded; and
determining a motion information predictor from the generated set of motion information predictors for the target image portion of the image to be decoded.