![Sample our Computer Science journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5928/TOC-Subject-Sample-2021-Computer-Science.jpg"})
Abstract
A new approach is proposed in this paper for image matching an invariant to translation, rotation, anisotropic scaling, and skewing. A compact image is acquired by principal component analysis and rescaling the image on the principal axes. However, the direction of the principal axis is susceptible to the skewing effect, which will bring about the failure of the orientation normalization. In addition, overall scale normalization is not taken into consideration. Hence, the central idea of this paper is to find the normalized orientation and overall scale for the compact image by virtue of the moment normalization. In consequence, the resultant image becomes thoroughly normalized and can be directly employed for matching by simple similarity metrics. Alternatively, for the purpose of fast matching, two 1‐D projected results are evaluated for matching. These are undoubtly faster than the conventional pixel‐based correlation.
Notes
Correspondence addressee