A study of a soft computing based method for 3D scenario reconstruction

Several recent works deal with 3D data in mobile robotic problems, e.g., mapping. Data comes from any kind of sensor (time of flight, Kinect or 3D lasers) that provide a huge amount of unorganized 3D data. In this paper we detail an efficient approach to build complete 3D models using a soft computing method, the Growing Neural Gas (GNG). As neural models deal easily with noise, imprecision, uncertainty or partial data, GNG provides better results than other approaches. The GNG obtained is then applied to a sequence. We present a comprehensive study on GNG parameters to ensure the best result at the lowest time cost. From this GNG structure, we propose to calculate planar patches and thus obtaining a fast method to compute the movement performed by a mobile robot by means of a 3D models registration algorithm. Final results of 3D mapping are also shown.

Figure optionsDownload as PowerPoint slideHighlights
► Our main goal is to perform six degrees of freedom pose registration in semi-structured environments.
► We use a method for modeling planar patches from 3D raw data, achieving: complexity reduction, outliers influence reduction.
► We propose the use of a Growing Neural Gas to represent fast and high quality 3D spaces.
► Modeling 3D scenes using GNGs produces a more detailed result and thus further computations are also improved.
► Our method is compare with classical ICP algorithm and results show the effectiveness of our method.