摘要
SAR图像具有很强的斑点噪声,使得单尺度边缘检测算子难以兼顾噪声抑制、检测边缘的完整性和定位的准确性。提出一种SAR图像的多尺度边缘检测方法。首先构造高斯多尺度边界检测算子,然后根据信号边界与噪声边界的小波变换模值跨尺度传递的不同特性,将不同尺度的检测算子检测的边缘相融合,提出由边缘传递、继承和生长构成的多尺度边缘关联融合算法。实验证明该方法可以有效克服斑点噪声的影响,融合的边界比较完整,定位准确。
Edge detection plays an important role in automatic process of SAR images. Single scale edge detector is difficult to take account of noise suppression, the integrity of edges and the veracity of position. A multiscale edge detect method is proposed. At first, multiscale Gaussian edge detector is constructed. According to transfer characteristic across scales of the wavelet modules of signal edge and noise edge, the property of edges in different scale is combined and a multiscale edge fusion algorithm is proposed consisting of edge transfer, edge inherit and edge growth. The result of experiment shows that this algorithm can get ride of the affect of noise,the edges fused have precise position and intact contour.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2004年第3期307-310,320,共5页
Systems Engineering and Electronics
关键词
合成孔径雷达
边缘检测
多尺度分析
synthetic aperture radar
edge detection
multiscale analysis
