PROJECT TITLE :

Graphics processing unit-accelerated multi-resolution exhaustive search algorithm for real-time keypoint descriptor matching in high-dimensional spaces

ABSTRACT:

Image keypoint descriptor matching is an important pre-processing task in various laptop vision applications. This study initial introduces an existing multi-resolution exhaustive search (MRES) algorithm combined with a multi-resolution candidate elimination technique to handle this issue efficiently. A graphics processing unit (GPU) acceleration style is then proposed to enhance its real-time performance. Suppose that a scale-invariant feature transform like algorithm is used to extract image keypoint descriptors of an input image, the MRES algorithm initial computes a multi-resolution table of every keypoint descriptor by using a L1-norm-primarily based dimension reduction approach. Next, a fast candidate elimination algorithm is used based mostly on the multi-resolution tables to get rid of all non-candidates from a candidate matching list by using a simple L1-norm computation. However, when the MRES algorithm was implemented on the central processing unit, the authors observed that the step of multi-resolution table building is not computationally economical, but it is terribly suitable for parallel implementation on the GPU. Therefore, this study presents a GPU acceleration technique for the MRES algorithm to attain better real-time performance. Experimental results validate the computational potency and matching accuracy of the proposed algorithm by comparing with three existing methods.