摘要 |
Embodiments relate to a method for encoding a video (V), executed by a source device (2), comprising encoding a current portion (Pc) of the video, by:
- determining (S2) a first observation (z k ) by comparing a first working state (x k ) of said current portion with said current portion (Pc),
- determining (S4) an updated first working state (x k ') in function of the first residual (z k ), and
- for at least one following current working state:
- determining (S6) a current working state (x k ) in function of the updated previous working state (x k-1 '),
- determining (S2) a current observation (z k ) by comparing the current working state (x k ) of said current portion with said current portion (Pc), and
- determining (S4) an updated current working state (x k ') in function of the current working state (x k ) and the current observation (z k ). |