A New Image Matching Algorithm Based on the Relative Position of Intensity
Xiaohang Hu, Shuangxia Pan, Han Bai · 2007
Illumination variations and noise pollution are the main factors that weaken most current image-matching algorithms. In order to get rid of these problems, a new image matching algorithm is proposed. It is based on every pixel's relative position of intensity (RPI) among its neighborhood instead of the intensity value of itself directly. A series of examples to demonstrate the algorithm and necessary results are also included. Experimental results reveal that the proposed RPI still can achieve correct matching under the condition of illumination variations, low signal noise ratio (SNR) and low fore-back ground contrast. This method serves as a general tool for high-precise image matching that are applicable to other vision problems such as 3-D reconstruction and satellite positioning.