A Novel Image Classification System Based on Evidence Probabilistic Transformation
Journal Title: International Journal of Research in Computer and Communication Technology - Year 2015, Vol 4, Issue 2
Abstract
This paper uses the evidence probabilistic transformation (EPT) for unsupervised image retrieval framework. The main advantages with EPT are substantially resolves the "take-them-or-leave-them" problem, gives a firmer epistemological basis for acquiring and for using in decisions. The proposed framework makes use of gray level co-occurrence matrix (GLCM) for images feature extraction. These features are used to provide a new axiomatic analysis and interpretation for images by using Dempster-Shafer belief functions. The mass functions for images information are combined under the normalized Dempster’s combination rule. The proposed model is characterized to extraction a set of accurate rules for images and their corresponding cluster based on probabilistic information in different format. Experiments show that the proposed model achieves a very good performance in terms of the precision, recall and F-measurement.
Authors and Affiliations
A. E. Amin
Keywords
Related Articles
- EP ID EP28142
- DOI -
- Views 257
- Downloads 1
Recent Issues and Its Quick Fixes for Various Data Repositories
The aim of this study was to assess the issues and their quick fixes of databases. But the database accuracy has a trickle down effect in most implementations. This paper proposes the overall guidance about the techni...
Security In Cloud Data Storage and Retrieval
Cloud data storage empowers clients to remotely store their information and appreciate the secured cloud applications without the complexity in equipment and programming administration. In spite of the fact that the...
A novel Cut Detection Technique in WSN
If some of the nodes fails in wireless sensor network it can get separated into multiple connected components which is called as “cut”. This paper concentrates on the problem of detecting cuts by the help remaining n...
A New approach for Cam-Shift Algorithm for Tracking a Moving Object
The main aim of this paper to track a target which plays a key component of video surveillance and monitoring systems. In this paper, we present an new approach to CamShift algorithm and kalman filter for tracking a...
The Safety Framework in Network Security utilizing Cryptographic Systems
Remote Sensor Networks comprise of independent sensor hubs appended to one or more base stations.As Wireless sensor systems proceed to grow, they get to be helpless against assaults and subsequently the requirement f...