基于TAPF算法的红外景象匹配特性研究  

Ternary amplitude-phase filter for infrared scene matching

作　　者：秦琴[1] 柴震海[1] 傅艳红[1] 施海燕[1] 王汝笠[1] 

机构地区：[1]中国科学院上海技术物理研究所,上海200083

出　　处：《红外与激光工程》2006年第4期490-494,共5页Infrared and Laser Engineering

摘　　要：红外景象匹配是自动目标识别领域的一个重要方向,如何在复杂的天气条件下提高红外景象匹配概率是研究的难点。在研究不同天气条件下的红外成像和传统匹配滤波器性能的基础上,提出了三值振幅-位相匹配滤波算法(TAPF),该算法结合了二值振幅信息与二值相位信息进行编码。对比灰度相关、二元纯位相匹配滤波和三值振幅-位相匹配滤波三种相关算法在不同天气下的结果,实验表明,提出的TAPF算法在相关面上的相关峰特性参数均优于其他两种算法,对恶劣的天气条件具有较强的适应性,是一种有效的红外景象匹配滤波算法。Infrared scene matching was an important subject in automotive target recognition field. How to improve the registration probability in complicated weather was a hard task. Ternary amplitudephase filter （TAPF） was proposed based on the infrared imaging in different weathers and the performance of traditional matching filter. This method combined the binary amplitude and the binary phase information together to realize TAPF. The performances of Grayscale, BPOF and TAPF in different weathers were compared. Experimental results show that TAPF＇s parameters in the correlation pin are better than the other two methods, and it has good adaptability to bad weathers. It is a valid matching filter in infrared scene matching field.

关 键 词：红外成像 匹配滤波器 目标识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]