Year-end Sale % OFF 
Abstract
Many of the point cloud processing techniques have their origin in image processing. But mathematical morphology, despite being one of the most used image processing techniques, has not yet been clearly adapted to point clouds. The aim of this work is to design the basic operations of mathematical morphology applicable to 3D point cloud data, without the need to transform point clouds to 2D or 3D images and avoiding the associated problems of resolution loss and orientation restrictions. The object shapes in images, based on pixel values, are assumed to be the existence or absence of points, therefore, morphological dilation and erosion operations are focused on the addition and removal of points according to the structuring element. The structuring element, in turn, is defined as a point cloud with characteristics of shape, size, orientation, point density, and one reference point. The designed method has been tested on point clouds artificially generated, acquired from real case studies, and the Stanford bunny model. The results show a robust behaviour against point density variations and consistent with image processing equivalent. The proposed method is easy and fast to implement, although the selection of a correct structuring element requires previous knowledge about the problem and the input point cloud. Besides, the proposed method solves well-known point cloud processing problems such as object detection, segmentation, and gap filling.
Highlights
Mathematical morphology (MM) is one of the main techniques of image processing
The results show a robust behaviour against point density variations and consistent with image processing equivalent
Unlike other techniques, such as Hough Transform (Borrmann et al, 2011), Region Growing (Vo et al, 2015), Connected Components (Tre vor et al, 2013), Histogram of Oriented Gradients (HOG) (Rahmani et al, 2014; Ren and Sudderth, 2016), Bag of Visual Words (BOVW) (Yu et al, 2016), Scale-Invariant Feature Transform (SIFT) (Jiang et al, 2018) or Convolutional Neural Networks (CNN) (Balado et al, 2020; Griffiths and Boehm, 2019), MM has not made a clear transition to the 3D point cloud environment
Summary
Mathematical morphology (MM) is one of the main techniques of image processing. It is based on object shape and set theory and was proposed and developed by Matheron & Serra in 1964 (Matheron and Serra, 2002). Its application in image processing is very useful for segmentation, refinement, feature extraction, and individu alization (Alegre et al, 2016; Serra and Soille, 2012) Unlike other techniques, such as Hough Transform (Borrmann et al, 2011), Region Growing (Vo et al, 2015), Connected Components (Tre vor et al, 2013), Histogram of Oriented Gradients (HOG) (Rahmani et al, 2014; Ren and Sudderth, 2016), Bag of Visual Words (BOVW) (Yu et al, 2016), Scale-Invariant Feature Transform (SIFT) (Jiang et al, 2018) or Convolutional Neural Networks (CNN) (Balado et al, 2020; Griffiths and Boehm, 2019), MM has not made a clear transition to the 3D point cloud environment. Skeleton of trees are obtained by implementing MM, through point cloud structured in voxels (Gorte and Pfeifer, 2004) or octrees (Bucksch and Wageningen, 2012)
Sign-in/Register to view highlights & summary 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 callMore From: ISPRS Journal of Photogrammetry and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
