SIFT-based feature extraction of high-definition core images automatic registration.
Wu Xiao · Journal of Sichuan University · 2009
This paper proposed Automatic Image Stitching technology to high-definition quality core scanning panoramic images. The key in automatic stitching is feature extraction and matching. Scale-invariant features transform algorithm with a telescopic, rotation, Affine invariant, and can resist a certain light and perspective changes. This paper proposed to apply scale-invariant features transform algorithm to image feature extraction, and used a Euclidean distance method of matching. Experimental results show that this algorithm greatly enhances the core image stitching level of automation and accuracy.