Abstract
This paper presents an approach of depth image segmentation based on the Evolving Principal Component Clustering (EPCC) method, which exploits data locality in an ordered data stream. The parameters of linear prototypes, which are used to describe different clusters, are estimated in a recursive manner. The main contribution of this work is the extension and application of the EPCC to 3D space for recursive and real-time detection of flat connected surfaces based on linear segments, which are all detected in an evolving way. To obtain optimal results when processing homogeneous surfaces, we introduced two-step filtering for outlier detection within a clustering framework and considered the noise model, which allowed for the compensation of characteristic uncertainties that are introduced into the measurements of depth sensors. The developed algorithm was compared with well-known methods for point cloud segmentation. The proposed approach achieves better segmentation results over longer distances for which the signal-to-noise ratio is low, without prior filtering of the data. On the given database, an average rate higher than 90% was obtained for successfully detected flat surfaces, which indicates high performance when processing huge point clouds in a non-iterative manner.
Highlights
With fast and reliable processing of sensor data, which enables high-level representation and perception of the observed environment, we can enable the development of emerging intelligent mobile robots that will be able of autonomous operation, such as fruit picking in an outdoor environment [7] or vacuuming interiors [8]
We focused on the literature related to our work and highlighted three commonly used methods for plane detection in a point cloud: approaches based on RANSAC, region growing (RG), and the Hough transform
We focus on 3D planar segmentation of the point cloud, namely the extraction of flat surfaces based on the aggregation of detected line segments
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Sensors 2021, 21, 4395 has been re-highlighted, suggesting the need to improve the process of eliminating such data These problems of depth data segmentation are addressed in our study, where we proposed a two-step filtering for the needs of reliable detection of outliers, which supports integration with the structure of online algorithms. We present an extension of the EPCC [28] method to 3D space for the detection of flat surfaces in ordered depth maps (point cloud) that enables the parallelization of data-stream processing due to its structure. Since depth map data are ordered, we can use data orderliness with the proposed approach to process a set of data in a single pass and to introduce a two-step filtering framework This makes the algorithm suitable for real-time applications.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
