- Title
- An unsupervised stochastic model for detection and identification of objects in textured color images using segmentation technique
- Creator
- Islam, Mofakharul; Watters, Paul
- Date
- 2009
- Type
- Text; Conference proceedings
- Identifier
- http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/45434
- Identifier
- vital:4306
- Identifier
-
https://doi.org/10.1007/978-90-481-9151-2_55
- Abstract
- The process of meaningful image object identification is the critical first step in the extraction of image information for computer vision and image understanding. The disjoint regions correspond to visually distinct objects in a scene. In this particular work, we investigate and propose a novel stochastic model based approach to implement a robust unsupervised color image content understanding technique that segments a color textured image into its constituent parts automatically and meaningfully. The aim of this work is to produce precise segmentation of different objects in a color image using color information, texture information and neighborhood relationships among neighboring image pixels in terms of their features using Markov Random Field (MRF) Model to get the maximum accuracy in segmentation. The evaluation of the results is done through comparison of the segmentation quality and accuracy with another similar existing method which demonstrates that the proposed approach outperforms the existing method by achieving better segmentation accuracy with faithful segmentation results.
- Publisher
- Online conference Springer
- Rights
- This metadata is freely available under a CCO license
- Subject
- Segmentation; Markov Random Field (MRF)
- Hits: 605
- Visitors: 599
- Downloads: 0
Thumbnail | File | Description | Size | Format |
---|