TY - GEN
T1 - A self-adaptable improved arithmetic based on wavelet transform in the image edge detection
AU - Jingchao, Zhang
AU - Songyan, Liu
AU - Guangpu, Hao
AU - Lingyun, Sun
AU - Jin, Yang
PY - 2009
Y1 - 2009
N2 - Compared with traditional edge detection methods, the method based on wavelet transform is a high precision way. The scale, however, impacts the effect of output imagine. The high frequency could be controlled when the scale is large, and the SNR of the image is relatively great. But the precision of location is lower. On the other hand, the location could be realized precisely when the scale is small, and the detail of the image could be extracted. But the ability of denoising is weak. A self-adaptable scale selection method is given in this paper. The edge detection of an image can be realized well by different properties of wavelet transform on different scales. Accordingly, the gauss filter is adopted to pre-treat the figure, and the self-adaptable multi-scale wavelet transform is introduced to process the after-pretreat figure. Result of experiment shows that a highly quality edge figure by smooth disposal of different scales is received.
AB - Compared with traditional edge detection methods, the method based on wavelet transform is a high precision way. The scale, however, impacts the effect of output imagine. The high frequency could be controlled when the scale is large, and the SNR of the image is relatively great. But the precision of location is lower. On the other hand, the location could be realized precisely when the scale is small, and the detail of the image could be extracted. But the ability of denoising is weak. A self-adaptable scale selection method is given in this paper. The edge detection of an image can be realized well by different properties of wavelet transform on different scales. Accordingly, the gauss filter is adopted to pre-treat the figure, and the self-adaptable multi-scale wavelet transform is introduced to process the after-pretreat figure. Result of experiment shows that a highly quality edge figure by smooth disposal of different scales is received.
KW - Edge detection
KW - Image processing
KW - Multi-scale analysis
KW - Self-adaptable filter
KW - Wavelet transform
UR - https://www.scopus.com/pages/publications/70449338612
U2 - 10.1109/ICMTMA.2009.130
DO - 10.1109/ICMTMA.2009.130
M3 - 会议稿件
AN - SCOPUS:70449338612
SN - 9780769535838
T3 - 2009 International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2009
SP - 414
EP - 417
BT - 2009 International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2009
T2 - 2009 International Conference on Measuring Technology and Mechatronics Automation, ICMTMA 2009
Y2 - 11 April 2009 through 12 April 2009
ER -