基于块运动类型与阈值相结合的运动估计算法

摘要

现有视频压缩法中大多都能够得到较好的压缩视频的效果,但其算法的运算过程往往存在着运算量大、复杂度高等问题.根据UMHexagonS算法的准确性与高效性,采用H.264 UMHexagonS快速运动估计算法,通过在其运算过程中加入阈值并且结合视频运动类型的自适应法,有效降低了搜索点数,提高了运动估计算法效率.通过在JM测试平台上进行了算法验证,验证了该算法能在保证编码性能的同时,快速、有效的减少算法的运算量与时间.与预期的结果相符,在无线视频监控领域中具有较好的发展前景.%Most of the existing video compression method can get nice video compression effect,however,the algorithm computation process with many problems such as need large amount of computation and high complexity and so on.For solve those problems,as UMHexagonS algorithms with high accuracy and efficiency,this paper using H.264 UMHexagonS fast motion estimation algorithm,by adding threshold to the computational process of H.264 UMHexagonS fast motion estimation algorithm and combined with the adaptive approach of video motion type,effectively reduce the search points number,improve the efficiency of the motion estimation algorithm.By algorithm verify on JM test platform,success verify the algorithm in this paper not only can guarantee the coding performance,but also can fast and effectively reduce the computation amount and algorithm time,consistent with the expected.Has a good development prospect in the field of wireless video surveillance.