一种粗到精分块金字塔预测连续排除算法  

A Coarse to Fine Subblock Pyramid Predictive Successive Elimination Algorithm

作　　者：王建东[1] 罗军[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院,湖南长沙410073

出　　处：《红外技术》2007年第10期598-602,共5页Infrared Technology

基　　金：国防预研基金项目

摘　　要：提出一种粗到精分块金字塔连续排除算法用于减少块匹配运动估计所需计算量。金字塔可以方便地从当前搜索窗的积分帧构建,连续的排除从金字塔顶层到底层逐级进行,将最小SAD值与金字塔提供的一系列越来越紧的界值相比较,可以跳过不必要的SAD计算。这样,算法在保证获得与全搜索算法相同精度的前提下,计算复杂度大大降低。仿真实验结果验证了本文算法与全搜索算法和几种快速搜索算法相结合的有效性。A coarse to fine subblock pyramid successive elimination algorithm was proposed to reduce the computational cost of block matching algorithms for motion estimation. Successive elimination was performed hierarchically from the top level to the bottom level of a pyramid that can be easily constructed form an integral frame of current search window, which can skip unnecessary SAD calculation by comparing minimum SAD with tighter and tighter decision boundaries obtained. Thus, the computational load can be saved while it can achieve the same estimation accuracy as the full search block matching algorithm. The efficiency of the proposed algorithm combined with full search algorithm and several fast search algorithms is verified by simulation.

关 键 词：金字塔 连续排除算法 多级连续排除算法 积分帧 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]