Fast and accurate plane segmentation in depth maps for indoor scenes

Hulík, Rostislav; Beran, Vitezslav; Španěl, Michal; Kršek, Přemysl; Smrž, Pavel

doi:10.1109/iros.2012.6385868

Cited by 33 publications

(15 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Oehler et al [2] performed Hough transformation and connected component analysis on the point cloud first as pre-segmentation and then applied RANSAC to refine each of the resulting "surfels" (2s per 640 × 480 points). Several algorithms [3]- [5] applied RANSAC on local regions of the point cloud (which decreases the data size considered in each RANSAC run so as to increase speed) and then grew the region from the locally found plane instance to the whole point cloud (0.2s [3] or 0.1s [4] per 640 × 480 points; 0.03s [5] per 320 × 240 points). Region-grow-based methods are another popular choice.…”

Section: B Related Workmentioning

confidence: 99%

Fast plane extraction in organized point clouds using agglomerative hierarchical clustering

Feng

Taguchi

Kamat

2014

2014 IEEE International Conference on Robotics and Automation (ICRA)

199

188

View full text Add to dashboard Cite

Abstract-Real-time plane extraction in 3D point clouds is crucial to many robotics applications. We present a novel algorithm for reliably detecting multiple planes in real time in organized point clouds obtained from devices such as Kinect sensors. By uniformly dividing such a point cloud into nonoverlapping groups of points in the image space, we first construct a graph whose node and edge represent a group of points and their neighborhood respectively. We then perform an agglomerative hierarchical clustering on this graph to systematically merge nodes belonging to the same plane until the plane fitting mean squared error exceeds a threshold. Finally we refine the extracted planes using pixel-wise region growing. Our experiments demonstrate that the proposed algorithm can reliably detect all major planes in the scene at a frame rate of more than 35Hz for 640×480 point clouds, which to the best of our knowledge is much faster than state-of-the-art algorithms.

show abstract

Section: B Related Workmentioning

confidence: 99%

Fast plane extraction in organized point clouds using agglomerative hierarchical clustering

Feng

Taguchi

Kamat

2014

2014 IEEE International Conference on Robotics and Automation (ICRA)

199

188

View full text Add to dashboard Cite

show abstract

“…Some RANSAC-based approaches have been extended to exploit organized structure, such as the approach of Biswas and Veloso, 13 which enables real-time performance. Hulik et al 14 proposed a tiled RANSAC for the ground plane search by taking into account only small areas of the scene in depth image. Because of its small random sample search, the tiled RANSAC can be used in real-time systems, which can reach a speed of multiple frames per second.…”

Section: -11mentioning

confidence: 99%

Fast plane segmentation with line primitives for RGB-D sensor

Zhang

Chen

Liu

2016

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

This article presents a fast and robust plane segmentation approach for RGB-D type sensor, which detects plane candidates by line segments extracted from 2-D scanline projected from row or column points. It neither requires high computation to calculate local normals for the entire point cloud as most of approaches do nor randomly chooses plane candidates such as RANSAC-like approaches. First, a line extraction algorithm is utilized to extract line segments. Second, the plane candidates are detected by estimating local normal of points lying on line segments. Finally, the plane having most inlier is recursively extracted from the plane candidates as the result plane. Experiments were conducted with different data sets and the segmentation performances were evaluated quantitatively and qualitatively. We demonstrated the efficiency and robustness of our proposed approach, especially in the none plane scenario, the approach needs little computational cost.

show abstract

“…State-of-the-art plane extracting methods from RGB-D images include RANSAC [Schnabel 2007] [Hulik 2012], clustering [Feng 2014] [Holz 2011] and region growing [Holz 2013] [Mostofi 2014] [Lee 2012]. All these methods only extract the planes from the 3D scenes, and don't interpret semantic information of these planes.…”

Section: Related Workmentioning

confidence: 99%

Parsing main structures of indoor scenes from single RGB-D image

Zhou

Pan

et al. 2016

JAMDSM

View full text Add to dashboard Cite

Parsing indoor scenes from RGB-D imagery is a significant and challenging task in image understanding and computer vision. Since the walls, floor and ceiling are the main structures of indoor scenes, we propose a simple and efficient method to find them from a single depth image. We use the prior knowledge of the walls, floor and ceiling and distance histogram to obtain their planes, and delimit their extents by ray-casting. Our method can also detect the open windows or doors and other planes which have large area in indoor scenes. Experimental results show that our method is more robust to find the main structures of the indoor scenes than the existing method.

show abstract

Fast and accurate plane segmentation in depth maps for indoor scenes

Cited by 33 publications

References 14 publications

Fast plane extraction in organized point clouds using agglomerative hierarchical clustering

Fast plane extraction in organized point clouds using agglomerative hierarchical clustering

Fast plane segmentation with line primitives for RGB-D sensor

Parsing main structures of indoor scenes from single RGB-D image

Contact Info

Product

Resources

About