基于分形的红外图象目标自动检测  

Fractal-Based Infrared Image Targets Automatic Detection

作　　者：卢福刚[1] 赵荣椿[1] 

机构地区：[1]西北工业大学计算机科学与工程系

出　　处：《信号处理》1999年第2期116-120,共5页Journal of Signal Processing

摘　　要：本文介绍了一种基于分形的复杂背景下红外图象目标自动检测方法，该方法以自然背景和人造目标在分形特征上的固有的差异为依据，利用了人造目标分形特征拟合误差大的特点，实现了复杂背景中红外图象目标的自动检测。实际检测的实验结果表明：该目标自动检测方法对环境的适应性强，是一种颇具发展前景的目标自动检测方法。In the paper, a fractal-based infrared image targets detecting method is present, which is based on the intrinsic difference in fractal featuures beetween natural background and man-made objects . It implement automatic infrared image targets detection from complex background by using high fractal fitting error of man-made objects. The practical detection results show that that this automatic targets detection method has a good adaptability to the environment and has a very good developing future.

关 键 词：分形 自动目标检测 红外图象 红外传感器 

分 类 号：TN216[电子电信—物理电子学] TP391.41[自动化与计算机技术—计算机应用技术]