A radial search method for fast nearest neighbor search on range images

Abstract

In this paper, we propose an efficient method for the problem of Nearest Neighbor Search (NNS) on 3D data provided in the form of range images. The proposed method exploits the organized structure of range images to speed up the neighborhood exploration by operating radially from the query point and terminating the search by evaluating adaptive stop conditions. Despite performing an approximate search, our method is able to yield results comparable to the exhaustive search in terms of accuracy of the retrieved neighbors. When tested against open source implementations of state-of-the-art NNS algorithms, radial search obtains better performance than the other algorithms in terms of speedup, while yielding the same level of accuracy. Additional experiments show how our algorithm improves the overall efficiency of a highly computational demanding application such as 3D keypoint detection and description.