A fast and fully distributed method for region-based image segmentation
- PDF / 1,920,003 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 64 Downloads / 177 Views
ORIGINAL RESEARCH PAPER
A fast and fully distributed method for region‑based image segmentation Fast distributed region-based image segmentation Smaine Mazouzi1 · Zahia Guessoum2 Received: 18 March 2019 / Accepted: 16 September 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract Distributed and parallel computing techniques allow fast image processing, namely when these techniques are applied at the low and the medium level of a vision system. In this paper, a collective and distributed method for image segmentation is introduced and evaluated. The method is modeled as a multi-agent system, where the agents aim to collectively produce a region-based segmentation. Each agent starts searching for an acceptable region seed by randomly jumping within the image. Next, it performs a region growing around its position. Thus, several agents find themselves within the same homogeneous region and are organized in a graph where two agents are connected if they are within the same region. So, a unifying of the labels in a same region is collaboratively performed by the agents themselves. The proposed method was experimented on real range images from the ABW dataset and the Object Segmentation Database (OSD) one, and the obtained results were compared to those of some well-referenced methods from the literature. The evaluation results show that the proposed method provides fast and accurate image segmentation, allowing it to be deployed for real-time vision systems. Keywords Image segmentation · Region growing · Pixel labeling · Distributed computing · Real-time computer vision systems
1 Introduction Image segmentation is a key low-level task in most of computer vision applications. It consists in partitioning the pixels of an image into homogeneous regions regarding a given criterium, mostly visual (color, texture, shape, etc.). In several cases, the segmentation process must be fast, suited for real-time applications, such as in remote-sensing and robotic vision. Methods for image segmentation can be split into two main categories: Edge-based methods, and region-based ones. For the methods in the first category, and by using some geometrical operators, the pixels that are located at the discontinuities in the image data are selected, chained and * Smaine Mazouzi s.mazouzi@univ‑skikda.dz 1
Computer Science Department, Skikda University 20 aout 1955, Skikda, Algeria
CReSTIC, University of Reims Champagne Ardenne, Reims Lip6, Sorbonne University, Paris, France
2
partitioned to produce the edges of the objects of the image [8, 18]. Edge-based methods are not time consuming, and can be used for real-time applications. However the quality of the results is usually not enough for further high-level processing. Indeed, such methods are highly sensitive to noise [3]. Moreover, they require intensive post-processing, which is necessary to provide a segmentation of the image into its constituent regions [24]. Unfortunately, such postprocessing tasks, mainly iterative using morphological operators,
Data Loading...
