Volume no :5, Issue no: 1, September (2016)

International Journal of Applied Mathematics and Machine Learning

RESEARCH ON THE TARGET AND INTERFERENCE ADHESION PROBLEM IN THE INFRARED IMAGE SEGMENTATION

Author's: Qi Xin, Pengfei Li, Shuaiqi Liu, Peng Geng and Weixuan Zhang
Pages: [15] - [24]
Received Date: August 6, 2016; Revised August 12, 2016
Submitted by: Jianqiang Gao.
DOI: http://dx.doi.org/10.18642/ijamml_7100121701

Abstract

The infrared images are more widely used in the military field and people’s daily life. There is adhesion between the targets and interferes because of the imaging effect is not ideal. In our paper, new infrared image segmentation based on the classic snake active contour model and mathematical morphology is proposed to resolve the interference adhesions problem. Firstly, image enhancement and noise smoothing method are used to enhance the infrared image. Secondly, snake active contour model is used to separate targets. Lastly, we use mathematical morphology method to separate the touched objects. The experiment show that satisfied object is obtained.

Keywords

infrared image segmentation, snake classical contour model, mathematical morphology.