Image matching using Gradient Orientation Selective Cross Correlation

Image matching has been an important topic in computer vision and image processing. A new approach named Gradient Orientation Selective Cross Correlation is proposed for image matching. In the new approach, a gradient orientation selectivity strategy is proposed to exclude the wrong points from correlation, especially for partial occlusion and some other ill-conditions. Experimental results are shown to demonstrate robustness, efficiency, and accuracy of the algorithm.