30 October 2009 An iris localization algorithm based on geometrical features of cow eyes
Author Affiliations +
Proceedings Volume 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis; 749517 (2009) https://doi.org/10.1117/12.832494
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China
Abstract
Abrupt vicious affairs on meat food safety show rapid expansion, which poses a threat to food safety and human health. In order to trace the flow of meat products in the food supply chain, the research on the individual identification of large animals based on iris recognition has very important practical significance. Iris localization is the key step in the iris recognition system, and this paper makes the cow iris as an example to propose two-step localization method on the basis of studying and analyzing lots of cow geometrical features. Firstly, threshold transform and Sobel edge detection operator are used to make the image binarization processing and realize coarse location, and the coordinates of boundary points are determined with the quadratic B-spline interpolation curves. Secondly, the cow eye iris image is located from the coarseness to fine through fitting circles. This paper proposes two methods to determine the parameters of the inner and outer boundary circles, including fitting concentric and non-concentric circle. Experimental results show that fitting non-concentric circle algorithm can locate iris region quickly and efficiently and has better accuracy and good robustness.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Menglu Zhang, Menglu Zhang, Lindu Zhao, Lindu Zhao, } "An iris localization algorithm based on geometrical features of cow eyes", Proc. SPIE 7495, MIPPR 2009: Automatic Target Recognition and Image Analysis, 749517 (30 October 2009); doi: 10.1117/12.832494; https://doi.org/10.1117/12.832494
PROCEEDINGS
8 PAGES


SHARE
Back to Top