Abstract
An automatic image matching based on Features from Accelerated Segment Test (FAST) corner detector and Local Binary Patterns (LBP) Descriptor is proposed in this paper. Firstly, constructing gauss pyramid of the reference image and the image to match, then detecting and extracting the FAST corner points of both images, and calculating the LBP descriptors of the corner points, following by searching match point pairs by K-D tree. Thirdly, using random sample consensus algorithm to remove error point pairs and then calculate the geometry transform coefficients. Lastly, executing the geometry transform to get the matching image. Experimental result proves that this method has higher matching accuracy, higher matching efficiency, good scale invariance and illumination invariance.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have