2024-03-29T01:38:57Z
https://u-ryukyu.repo.nii.ac.jp/oai
oai:u-ryukyu.repo.nii.ac.jp:02001973
2022-02-14T21:19:52Z
1642837622505:1642837855274:1642837877045
1642838403551:1642838406845
Edge Detection by Pattern Matching Based on Principal Component Analysis
Zeng, Xiang-Yan
Chen, Yen-Wei
Nakao, Zensho
陳, 延偉
仲尾, 善勝
boundary detection
principal component analysis
edge pattern
image patches
In this paper, we have proposed a new method for edge detection by pattern matching. There exist two main problems for the edge detection methods by gradient operators: (1) gradients in the templates are fixed and not related with those in real images; (2) the results are edge enhanced images and need to be transformed into binary ones by setting an appropriate threshold. In our method, the pattern templates are obtained by learning the principal components of image patches and edges are represented directly by the class of the matched templates. The effectiveness of the proposed method is demonstrated by experimental results
紀要論文
http://purl.org/coar/resource_type/c_6501
琉球大学工学部
NA
http://hdl.handle.net/20.500.12000/1955
0389-102X
AN0025048X
琉球大学工学部紀要
61
68
65
eng
publisher
open access