摘要
针对图像序列中多个运动弱小目标的检测问题,提出了一种用于基于小波变换与帧间差分相结合的检测方法。算法首先对原始图像进行小波变换,利用小波变换增强了目标与背景之间的差别,然后使用帧间差分方法对小波变换后的水平分量和垂直分量进行运算,计算出多个运动的弱小目标。与传统的帧差分方法相比,该方法利用小波变换增强了目标与背景之间的差别,帧间差分运算的区域小,速度快。实验结果表明,使用该方法可以对信噪比为2.9的目标正确提取,减小了虚警率,抗噪声干扰能力强。
A algorithm based on wavelet transform and frame difference method is proposed for multi moving small target in series im- ages. This method will wavelet transform on image in order to suppress background and enhance targets, then compute on horizon and vertical by frame difference method in order to compute the target. Compared with the traditional frame difference method, this method enhances the targets, the size of frame difference method is more smaller and the speed is faster. Experiment result show t...
出处
《微计算机信息》
北大核心
2008年第10期277-278,共2页
Control & Automation
关键词
小波变换
目标检测
低信噪比
多目标
图像序列
Wavelet transform
Target detecting
Low SNR
Multi- target
image sequence