Planar segmentation of range images

[摘要] ENGLISH ABSTRACT: Range images are images that store at each pixel the distance between the sensor and a particularpoint in the observed scene, instead of the colour information. They provide a convenient storageformat for 3-D point cloud information captured from a single point of view. Range imagesegmentation is the process of grouping the pixels of a range image into regions of points thatbelong to the same surface. Segmentations are useful for many applications that require higherlevelinformation, and with range images they also represent a significant step towards completescene reconstruction.This study considers the segmentation of range images into planar surfaces. It discusses thetheory and also implements and evaluates some current approaches found in the literature. Thestudy then develops a new approach based on the theory of graph cut optimization which hasbeen successfully applied to various other image processing tasks but, according to a search ofthe literature, has otherwise not been used to attempt segmenting range images.This new approach is notable for its strong guarantees in optimizing a specific energy functionwhich has a rigorous theoretical underpinning for handling noise in images. It proves to be veryrobust to noise and also different values of the few parameters that need to be trained. Resultsare evaluated in a quantitative manner using a standard evaluation framework and datasets thatallow us to compare against various other approaches found in the literature. We find that ourapproach delivers results that are competitive when compared to the current state-of-the-art,and can easily be applied to images captured with different techniques that present varying noiseand processing challenges.

[发布机构] Stellenbosch University