[其它]Curvature Scale Space Corner Detector

Abstract 6YLj^w] %  
Corners play an important role in object identification +]A:M6P:{v  
methods used in machine vision and image processing 3kIN~/<R+7  
systems. Single-scale feature detection finds it hard to OgQV;at  
detect both fine and coarse features at the same time. On ?U5{Wa85D  
the other hand, multi-scale feature detection is inherently 6?mibvK  
able to solve this problem. This paper proposes an AOWmzu{zw  
improved multi-scale corner detector with dynamic region zRl3KjET  
of support, which is based on Curvature Scale Space (CSS) '}JhzKNj  
technique. The proposed detector first uses an adaptive [.8BTj1%  
local curvature threshold instead of a single global %C'?@,7C  
threshold as in the original and enhanced CSS methods. k)= X}=w  
Second, the angles of corner candidates are checked in a ?'Xj g#}<  
dynamic region of support for eliminating falsely detected W*4-.*U8a  
corners. The proposed method has been evaluated over a t?ZI".>  
number of images and compared with some popular #!qm ZN  
corner detectors. The results showed that the proposed .. #ASz;$P  
U;V7 u/{  

未注册仅能浏览部分内容，查看全部内容及附件请先 登录 或 注册