A model-based range image segmentation algorithm using a novel robust estimator
Hanzi Wang, David Suter · 2003
This paper presents a novel range image segmentation algorithm based on a newly proposed robust estimator: Adaptive Scale Sample Consensus (ASSC) [28]. The proposed algorithm is a model-based top-down technique and directly extracts the required primitives (models) from the raw images. Compared with current popular methods (region-based and edge-based methods), the algorithm is very robust to noisy or occluded data due to the adoption of the novel robust estimator ASSC. Using a hierarchical implementation, the proposed method is computationally efficient. Experimental results on real range images show that the proposed algorithm is attractive when compared with other state-of-the-art segmentation methods.