摘要
在用光电混合联合变换相关器对运动目标进行检测的过程中,为了在提高运算速度的前提下解决运动目标在低对比度环境中所导致的低识别率问题,提出一种基于提升小波的边缘检测算法,该算法能在将目标边缘准确地从低对比度环境中提取的同时,将传统小波变换的计算量减少一倍,显著提高运算速度,满足运动目标检测在处理速度上的需求。将该算法应用于光学相关检测,并结合瞬态模板方法可改善目标的识别率,提高在低对比度环境下运动目标检测的效率。对大量具有低对比度特点的运动序列进行光学相关实验,验证了算法对低对比度环境下运动目标检测的可行性。
In the process of detecting moving target with hybrid optoelectronic joint transform correlator （ HOJTC ）, in order to solve the low recognition ratio problem of moving target under low contrast environment and improve the calculation speed, this paper offers an edge extraction algorithm based on lifting wavelet, which can extract target edge from low contrast background exactly,reduce the calculation task of traditional wavelet transform by half, significantly enhance the calculation speed, and meet the processing speed requirement in moving target detection. This algorithm is applied to optical correlation detection, and combined with temporal template method the target recognition ratio is improved and the moving target detection efficiency under low contrast environment is enhanced. The optical correlation experiments on a large number of moving sequences with low contrast characteristic were carried out. Experiment results verify the feasibility of the proposed algorithm in moving target detection under low contrast environment.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2013年第2期319-325,共7页
Chinese Journal of Scientific Instrument
基金
总装备部预研局十一五基金(51******05)资助项目
关键词
光学相关检测
低对比度环境
运动目标
提升小波
边缘检测
瞬态模板
optical correlation detection
low contrast environment
moving target
lifting wavelet
edge detection
temporal template
