A novel macro - block layer rate control algorithm in low bit - rate 一種低碼率下的新型宏塊級碼率控制算法
With a view to dct that costs the longer computing time , fast dct is proposed , and the method of prejudgment of an macro - block of all zero coefficient of dct is adopted in this paper , which makes about 40 percent dct computation reduced , consequently , the encoding speed of h . 263 encoder can be further increased 263編碼器的編碼速率;對運算量較大的dct給出了快速算法,并對全零dct系數塊采用提前判斷的方法,從而省去了約40的dct運算,進一步提高了h 263編碼器的編碼速率。
Firstly , we directly use the motion vectors of macro - blocks defined in mpeg - i / ii compressing standards and filter the immobile macro - blocks . then , we build a skin color model in ycbcr color space using the convergent property of skin color , and we present the gaussian model skin recognition method and positive - negative look - up table method in details . and we analyze the texture of skin after wavelet transform and present a bayesian method based texture recognition method and a high texture filtering method 根據皮膚的運動性,首先直接利用mpeg -中的壓縮標準中有關宏塊運動預測的方法,提取宏塊的運動矢量,將沒有運動的宏塊過濾掉;然后,利用皮膚顏色的聚合性,在ycbcr顏色空間建立了皮膚的顏色模型,并分別闡述了基于高斯分布模型的皮膚檢測法和正反概率表方法;最后,通過對皮膚進行小波變換后的紋理進行統計后,發現有效的利用皮膚紋理特征,可以比較有效的過濾掉那些具有類似于皮膚顏色的背景,分別闡述了基于貝葉斯方法的紋理檢測方法和高紋理過濾法。
For video codec , after analyzing and researching the newest standard h . 264 , an efficiency algorithm was proposed in this paper using self - feature , spatial correlation and temporal correlation to judge the encoding mode of one macro - block . through this method , the best prediction mode can be obtain before encoding , then , the coding time can be reduced dramatically with a little expense of bit - rate . for video transmission , in this paper , a transmit method was proposed using tcp / ip protocol under the environment of local networks 首先,在分析及研究最新的視頻編碼標準h . 264的基礎上,利用宏塊自身的特點和其時間、空間的相關性來判斷宏塊應該采用的編碼模式,從而提前得到該宏塊的最佳編碼模式,新方法有效地克服了原標準算法中窮舉式的復雜算法,在壓縮碼流只有少許增加的情況下,大幅度地提高了幀內及幀間編碼的速度。