Edge detection using matched filter

For target tracking, automatic target recognition and detection applications, they value edge detection sensibility, precision and location accuracy rather than other criterions. Aimed at these three criterions, this paper presents an edge detection algorithm based on matched filter. Firstly, a matc...

Full description

Saved in:
Bibliographic Details
Main Authors: Luo Hai-bo, Jiao An-bo, Xu Ling-yun, Shao Chun-yan
Format: Conference Proceeding
Language:English
Subjects:
edge detection
Feature extraction
Filtering theory
Finite impulse response filters
Image edge detection
local adaptive threshold
matched filter
Matched filters
Mathematical model
Noise
noise suppression
real-time application
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:For target tracking, automatic target recognition and detection applications, they value edge detection sensibility, precision and location accuracy rather than other criterions. Aimed at these three criterions, this paper presents an edge detection algorithm based on matched filter. Firstly, a matched filter was designed by analyzing the edge model for natural image, then the edge response was computed using the designed matched filter; secondly, the lower and discontinuous filtered responses were further suppressed using a dedicated one-dimension filter; finally, the edge image was obtained by binarizing the edge response with a local adaptive threshold. Experimental results illustrate that the proposed algorithm has more improvement than the Sobel and Canny operators in detection sensibility, precision and location accuracy. Moreover, the algorithm can be implemented with parallel pipeline using FPGA, so it is also rather suitable for real-time applications.
ISSN:1948-9439
1948-9447
DOI:10.1109/CCDC.2015.7162087