Abstract
In order to quickly and exactly extract the subpixel image edge, a new method of subpixel image edge based on Fourier-Mellin moment is proposed. Firstly all possible edge points were detected quickly with using Fourier-Mellin moment operator with mask size of 3×3, secondly a Fourier-Mellin moment operator with mask size of 7×7 was used to relocate the edge with subpixel accuracy. Experiment results show that the method has good subpixel accuracy, and its processing time is 1.53s. Therefore, the method has advantages of both short running time and high accuracy.
Original language | English |
---|---|
Title of host publication | 2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 - Proceedings |
Pages | 1293-1296 |
Number of pages | 4 |
DOIs | |
Publication status | Published - 2011 |
Event | 2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 - XianNing, China Duration: 16 Apr 2011 → 18 Apr 2011 |
Publication series
Name | 2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 - Proceedings |
---|
Conference
Conference | 2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 |
---|---|
Country/Territory | China |
City | XianNing |
Period | 16/04/11 → 18/04/11 |
Keywords
- OFMM
- edge detection
- model effect
- subpixel edge
Fingerprint
Dive into the research topics of 'Fast subpixel edge extraction based on two Fourier-Mellin moments with different mask sizes'. Together they form a unique fingerprint.Cite this
Zeng, T., Shi, G., Zheng, H., & Wang, Z. (2011). Fast subpixel edge extraction based on two Fourier-Mellin moments with different mask sizes. In 2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 - Proceedings (pp. 1293-1296). Article 5769028 (2011 International Conference on Consumer Electronics, Communications and Networks, CECNet 2011 - Proceedings). https://doi.org/10.1109/CECNET.2011.5769028