介绍了小波变换及其应用于图像边缘多尺度检测的原理，并给出了构造小波函数的方法和相对应的滤波器系数。用一幅图像进行的仿真实验表明，基于小波变换进行图像的多尺度边缘检测相对于传统的Roberts算子、Sobel算子和Laplacian算子检测图像边缘是一瓣较好的方法，而且对随机噪声也有较强的鲁棒性。为此进一步提出从多个目标中快速跟踪和识别指定目标的思想。
Abstract
The paper describes wavelet transform and the principle of its application in multiscale edge detection. The paper aslo provides the method of construction of the wavelet function. Simulation tests show that multiscale edge detection of images using wavelet transform is better when compared with the conventional Roberts operator, Sobel operator and the Laplacian operator, and that it exhibits a better robustness. It is further pointed out that the idea is well applicable in the fast recognition of targets.
关键词
小波变换 /
图像处理 /
边缘检测 /
目标识别
{{custom_keyword}} /
Key words
wavelet transform /
image processing /
edge detection /
object recognition
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Rioul O , Vetterti M. Wavelets and signal processing. IEEE Signal Processing Magazine, 1991
[2] Semmes S. Nonlinear fourier analysis. Bulletin of the AMS, 1S89, 20
[3] 焦李成，保铮.子波理论与应用：进展与展望.电子学报，1993 (7)
[4] Rosenfeld A, Thurston. Edge and curve detection for visual scene analysis. IEEE Trans. Com- put, 1970 (20): 562-575
[5] Marr D, Hildreth E. Theory of edge detection. Proc. Royal Soc.，London ： 1980： 187-217
[6] Pitas I, Venetsauopoulos A N. Edge detectors based on nonlinear filter. IEEE PAMI, 1986 (8)： 538-547
[7] Vincent T, Poggio T A. On edge detection. IEEE PAMI 1986 (2)： 147-163
[8] Mallat S, Zhong S. Characterization of signals from multiscale edges. IEEE PAMI, 1992 (14). 710-731
[9] Mallat S, Hwang W L. Singularity detection and processing with wavelets. IEEE Trans, on Information Theory, 1992, 38: 617-634
[10] Mallat S. A theory for multiresolution signal decomposition. IEEE PAMI, 1989 (7)： 674-693
[11] Mallat S. Zero-crossings of a wavelet transform. IEEE Trans, on Information Theory , 1994(4): 1019-1033
[12] 秦前淸，杨宗凯?实用小波分析.西安电子科技大学出版社，1994
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
