摘要
提出了一种基于小波与分形的红外小目标自动检测算法 ,首先对原始图像进行小波分解 ,计算其低频概貌图像的分形参数 ,利用在不同纹理图像分形参数变化的特性 ,检测出水天线的位置 ,在水天线附近利用红外目标的灰度特性进行精细搜索 ,最后确定目标所在的位置。实验表明该算法能有效地消除云层、海浪等杂波干扰 。
An algorithm for the automatic detection of infrared small sea target based on wavelet and fractal is proposed.Firstly,the original image is decomposed to subimages.Secondly,by computing the fractal parameter of the low frequency subimage and using the variety of fractal parameter in different texture images,we can confirm the position of skyline.Lastly,considering infrared targets usually have high value of intensity,we can find the target by carefully search near the skyline.Experiments show that this algorithm is effective to eliminate the clutters of clouds and offing,and is suitable for the detection of remote infrared small target under the complicated background of sea and sky.
出处
《激光与红外》
CAS
CSCD
北大核心
2004年第6期481-483,共3页
Laser & Infrared
基金
中科院先进制造基地创新课题基金资助项目(F0 10 40 8)。
关键词
小波
分形
图像处理
目标检测
wavelet
fractal
image processing
target detection
