提出了一種近似估計整幀圖像最小量化失真度的模型和新的率失真優(yōu)化的實時幀內(nèi)更新準則，可使編碼器直接計算出編碼一幀圖像的最小量化失真度，通過評估宏塊幀內(nèi)編碼所增加的總量化失真度和減小的擴散失真度，以決定是否需要對此宏塊進行幀內(nèi)更新。經(jīng)與傳統(tǒng)的基于率失真的幀內(nèi)更新相比和實驗結(jié)果表明，更新算法在具有更小計算復雜度的同時，在大多數(shù)情況下可以取得更好的峰值信噪比值。
關 鍵 詞 視頻編碼; 差錯恢復; 幀內(nèi)更新; 率失真優(yōu)化

Abstract A rate-distortion computation model is established for approximately estimating the total quantization distortion of one frame, and a new real time Intra-update criterion is proposed. Through comparing the increased quantization distortion and reduced propagation distortion by intra-updating, the encoder can quickly decide which macroblock need to be intra-updated. Simulation results show that the performance of proposed algorithm, which reduces the computational complexity significantly, achieves the same or even better gains over traditional rate-distortion based Intra-update methods.
Key words video coding; error resilience; intra-update; rate-distortion optimization