摘要
针对目前图象边界检测中存在的难点,本文系统阐述了主动型边界检测的思想,试图将边界检测由传统的信号处理过程转化为信息的能动提取过程,并介绍了在这一思想指导下进行的一些初步研究;研究结果表明,边界的主动提取方法给边界检测过程赋予了主动性,从而能够综合利用、权衡各种信息,提高边界检测的精度和效率。
As to the difficulties of the current frame of edge detection in early vision, the idea of ac-tive edge detection is described in this paper. According to this idea , the edge detection which has beenconsidered as a traditional siginal processing procedure is now attempted to be transfered to an active in-formation extraction procedure controlled by the feedbacks depending on current states of the data inter-pretations and the goals of the tasks in each stepe. Some initial researches motivated by the idea are alsointroduced. Research results show that the active edge extraction methed gives activity to edse detec-tion. Thus,many kinds of information can be utilized and analysezed to improve the accuracy and effi-ciency.
出处
《东南大学学报(自然科学版)》
EI
CAS
CSCD
1995年第4期47-53,共7页
Journal of Southeast University:Natural Science Edition
基金
江苏省"八五"重点攻关项目
关键词
机器视觉
图像处理
特征抽取
边界检测
computet vision
image processing
feature extraction
early vision
edge detection