Target Detection Algorithm Based on Fractal Features of Variance

Fractal dimension has been widly used for man-made targets detection because it can reflects an image's roughness. However, it always performs not very well because the other feature- intercept which reflects the the speed of surface change is neglected. Aimed to this problem, an improved algor...

Full description

Saved in:
Bibliographic Details
Main Authors: Dong, Wen-Ming, Wu, Le-Hua, Jiang, De-Lei, Han, Chao
Format: Conference Proceeding
Language:English
Subjects:
Brownian motion
Change detection algorithms
Chaotic communication
Detection algorithms
Fractals
Geometry
Object detection
Rough surfaces
Surface fitting
Surface roughness
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Fractal dimension has been widly used for man-made targets detection because it can reflects an image's roughness. However, it always performs not very well because the other feature- intercept which reflects the the speed of surface change is neglected. Aimed to this problem, an improved algorithm about how to calculate the fractal dimension based on variance property of fractional brownian motion is presented and utilized with intercept feature together. And then,a detection algorithm is proposed. Experiments show that the method has good adaptability and improved the efficiency, whether the background or the objects has changed or not.
DOI:10.1109/IWISA.2009.5073157