The invention, provided for an implementation, in a video encoder with base layer coding means and enhancement layer coding means, relates to a method of coding the video objects (VOs) of a sequence according to the following steps : segmentation of the sequence, and coding operation of the texture and shape of said VOs. According to a preferred embodiment, the texture coding operation itself comprises motion compensated prediction operations, during which the temporal references of the enhancement layer VO planes (VOPs) of type P or B are selected only on a temporal distance criterion, without any consideration of the layer they belong to. |