摘要
利用快速傅里叶变换FFT将图像信号从空间域转换到频域进行分析,使快速卷积、目标识别等许多算法易于实现;然后根据图像信号的灰度结构特征和频谱分布,用Butterworth带通滤波器和二维维纳滤波器进行滤波处理,去除图像信号中的低频干扰和噪声信号;再利用傅里叶反变换将信号还原。结果显示,处理后的模拟远程高空卫星照片轮廓清晰可见。
A lot of algorithms such as fast convolution and targets identification can be easily performed through converting image signal from spatial domain to frequency domain and by means of Fast Fourier Transform (FFT). Filtering processing for eliminating low-frequency disturbance and noise signal is carried out by Butterworth band pass filter and 2-D Wiener filter according to gray scale structural behavior and spectrum distribution of the image signal. The signal can be restored by Fourier inversion. The results show that after processing the photo of the simulated remote high-altitude secondary planet can be seen clearly.
出处
《光电工程》
CAS
CSCD
北大核心
2004年第B12期1-3,7,共4页
Opto-Electronic Engineering
关键词
快速傅里叶变换
数字图像处理
卫星照片
目标识别
Fast Fourier transforms
Digital image processing
Satellite photo
Target recognition