Segmentation of Natural Images and Retrievals Based on the Mixture of Pearson Type III Distributions
P. Chandra Sekhar1, K Srinivasa Rao2, P Srinivasa Rao3
1Dr P Chandra Sekhar, Department of IT, GITAM deemed to be University, Visakhapatnam, India.
2Prof. K Srinivasa Rao, Department of Statistics, Andhra University, Visakhapatnam, India.
3Prof. P Srinivasa Rao, Department of CSSE, Andhra University, Visakhapatnam, India.
Manuscript received on 29 May 2019 | Revised Manuscript received on 05 June 2019 | Manuscript published on 30 June 2019 | PP: 2038-2042 | Volume-8 Issue-8, June 2019 | Retrieval Number: H6818068819/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In the real world, the image retrievals and image analysis are most important for computer vision and security, surveillances video processing and remote sensing. For authentication and identification an image regional variation are more important. The segmentation of the image plays a vital part in identification of image regions. Among different segmentation techniques of the image, segmentation methods based on a model are prominent and provide accurate results. it is reasonable to consider the probability model, which closely matches with the physical features of image region for describing a suitable model. In the present paper, a novel and new segmentation methods of the image is carried using Type III Pearson system of distributions. In the experimentation one has to assume the image is exemplify with a K-component concoction of Pearson Type III distribution. The EM(Expectation Maximization) algorithm is used to predict the variables of the model. Three images of the real world are arbitrarily chosen from Berkeley database through experimentation. The computed values of VOI, GCE and PRI revealed that proposed method provide more accurate results to same images in which the image regions are left skewed and having long upper tiles. Through image eminence metrics the performance of image retrievals with proposed method is also studied and found that this method performs well then segmentation method based on GMM (Gaussian Mixture Model).
Keywords: Expectation Maximization (EM), Image Segmentation, Pearson Type III, Image Eminence Metrics.
Scope of the Article: Natural Language Processing