| 摘要 |
Deriving and selectively enabling a set of video coding modes for images or image portions in an image sequence based on coding mode information obtained from previously encoded image(s) of the sequence. Previously encoded images may form a dependency set based on image similarity such as temporal stamp (display order), coding stamp (coding order), temporal layer index, quantization parameters, Lagrangian parameters, or image type. A distance function based on a weighted characteristics vector may be calculated. Coding mode decisions may be based on analysed statistics of the images including best mode, rate distortion penalty and cost. Images may belong to layers of scalable video of the High Efficiency Video Coding HEVC standard. Video encoding and decoding methods and devices are disclosed. Multi-layer video coding may involve several versions of the same content at different resolutions; this method of adaptive video coding evaluates each coding mode for optimal mode selection and has improved speed and reduced bandwidth requirements. |
