摘要
针对战场环境和打击目标类型的不同,研究了图像目标自动识别和快速筛选技术,提出了大气云层背景下低信噪比弱信号小目标的检测截获、多目标边跟踪、边识别技术方案;然后针对海天图像的特点,分析了海杂波条件下两种海天背景的自适应分割算法(基于统计直方图的分割算法和基于梯度均值的分割算法);最后讨论了典型地面目标(阵列目标、港口目标等)的自动识别方法.仿真试验和外场测试试验验证了算法的有效性和实用性.
Aimed at battlefield environment and different kinds of struck targets,image target automatic recognition and fast filtering techniques are discussed,respectively.Firstly,the method for small target detection and recognition in cloud clutter background is given;Then based on the features of sea-space images,two adaptive segmentation algorithms for sea-space background targets(segmentation algorithm based on statistic histogram and segmentation algorithm based on gradient average)are analyzed;finally automatic recognition algorithms for typical ground targets such as array targets and port targets are brought forward.The effectiveness and practicability of the above algorithms are validated by simulation data and outfield tests.
出处
《红外与激光工程》
EI
CSCD
北大核心
2006年第z4期225-231,共7页
Infrared and Laser Engineering
关键词
自动目标识别
复杂背景
图像识别
Automatic target recognition
Complicated background
Image recognition
