Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
IMAGE REGION SELECTION METHOD AND SYSTEM FAVORABLE FOR FUZZY KERNEL ESTIMATION
Document Type and Number:
WIPO Patent Application WO/2019/010932
Kind Code:
A1
Abstract:
An image region selection method and system favorable for fuzzy kernel estimation. The implementation of the method comprises: calculating a relative total variation value of each pixel point in a fuzzy image and obtaining a relative total variation mapping graph of the fuzzy image; setting a threshold to determine whether each pixel point in the image is a boundary pixel point; then performing sampling on the fuzzy image and the relative total variation mapping graph of the fuzzy image so as to obtain a series of image blocks; and finally, collecting statistics on the number of boundary pixel points in each image block and selecting an image region favorable for fuzzy kernel estimation. The method effectively solves the problems, such as excessive dependency on operator experience and low efficiency, in an existing region selection method; an image region favorable for fuzzy kernel estimation is automatically selected; and the method is suitable for the application scenario of fuzzy kernel estimation in an image defuzzification algorithm.

Inventors:
SANG NONG (CN)
LI LERENHAN (CN)
YAN HAO (CN)
GAO CHANGXIN (CN)
YAN LUXIN (CN)
ZHANG SHIWEI (CN)
PI ZHIXIONG (CN)
Application Number:
PCT/CN2018/071692
Publication Date:
January 17, 2019
Filing Date:
January 08, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
UNIV HUAZHONG SCIENCE TECH (CN)
International Classes:
G06V10/44; G06T5/00; G06V10/50; G06V10/22
Foreign References:
CN107403414A2017-11-28
CN102436639A2012-05-02
CN102663718A2012-09-12
CN105096316A2015-11-25
CN102254309A2011-11-23
Attorney, Agent or Firm:
HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY PATENT AGENCY CENTER (CN)
Download PDF: