论文部分内容阅读
为了降低低码率视频编码算法的块效应与环效应，提出了一种基于自适应分区余弦变换／子波变换和重叠运动补偿的视频编码算法。首先给出一种空－时相关自适应运动估计算法，有效降低运算复杂性；然后通过分层游动窗块区域形成算法检测出帧间预测误差场的高能量区域，对大区域进行子波变换和量化，对零散区域进行余弦变换和量化；根据人类视觉系统（ＨＶＳ）的空－频和时－频特性，给出基于ＨＶＳ的量化模型，以期在信噪比约束下获得最好的主观测试质量．模拟结果表明，对于ＣＩＦ格式图像，在低码率和甚低码率下仍能得到满意的图像质量；算法向下兼容ＩＴＵ的Ｈ．２６３标准，并可以应用于低码率视频会议系统中．
In order to reduce the block effect and ring effect of low bit rate video coding algorithm, a video coding algorithm based on adaptive partition cosine transform / wavelet transform and overlapped motion compensation is proposed. Firstly, a space-time correlation adaptive motion estimation algorithm is proposed to reduce the computational complexity effectively. Then the high-energy region of the inter prediction error field is detected by the layered moving window region formation algorithm, Transform and quantify the cosine transform and quantize the discrete regions. According to the space-frequency and time-frequency characteristics of the human visual system (HVS), the HVS-based quantization model is given to obtain the best subjective Test the quality. The simulation results show that, for CIF format images, satisfactory image quality can still be obtained at low bit rate and very low bit rate; the algorithm is backward compatible with ITU H 263 standard, and can be applied to low bit rate video conference system.