The Quick Method for Image Registration Combined with Information Entropy and Cross-Correlation Matrix

Abstract:

The current image registration technique based on gray-level information has shortcomings on time and amount of computation brought by the whole image registration, this paper proposes a quick registration method, which makes a combination of image information entropy and cross-correlation matrix: First, segmenting the target image into blocks and calculating to get the maximum entropy image block, then using it as a template to calculate the cross-correlation matrix with floating image; Second, making the point where the maximum cross-correlation value ​​is as the upper left corner, grasping a same-size block with template on floating image; Finally, obtaining registration parameters through calculation for these two blocks to achieve the purpose of registration. Experimental results show that this method has less computational complexity with the similar registration results, and takes less time. It’s effective and feasible.

• Periodical: Advanced Materials Research (Volumes 971-973)
• Pages: 1812-1815
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.971-973.1812
• Online since: June 2014
• Authors: Yan Hai Wu*, Fang Ni Zhang
• Keywords: Block, Cross-Correlation Matrix, Image Registration, Information Entropy, Normalized Cross-Correlation Function, Real-Time

References

[1] Chen Xianyi. Image registration technique and MATLAB programming [M]. Beijing: electronic industry press, 2009, 1-234.

[2] Wu Zepeng, Guo Lingling, Zhu Mingchao. Combined with image information entropy and feature points of image registration method [J]. Infrared and laser engineering, 2013, 42（10）：2846-2852.

[3] Yamamura Y, Kim H, Yamamoto A. A method for image registration by maximization of mutual information[C]/SICE-ICASE, 2006. International Joint Conference. IEEE, 2006: 1469-1472.

DOI: 10.1109/sice.2006.315301

[4] Chen Qingfang. Based on mutual information and optimization algorithm of multimode image registration [D]. Jiangnan university, (2011).

[5] Song Ying, Chen Ke, Lin Jiangli. The edge detection algorithm based on image block [J]. Computer engineering, 2010, 36（14）: 196-200.

[6] Xu Suhui, Mu Xiaodong, Ke Bing. Contain large flat area high resolution remote sensing image fast registration [J]. Journal of computer applications, 2013, 33（SI）：182-185.

[7] Heather J P, Smith M I. Multimodal image registration with applications to image fusion[C]/Information Fusion, 2005 8th International Conference on. IEEE, 2005, 1: 8 pp.

DOI: 10.1109/icif.2005.1591879