Analysis of Random Projection & Live –Wire For Texture Extraction & Segmentation

Abstract

Automatic image segmentation requires a certain degree of depth of work combining image clustering, feature extraction, statistical analysis and iterative feedback. Researchers have been using algorithms like Principal component analysis (PCA) combined with advanced feature extraction techniques like gray level co-occurrence integrated algorithm (GLCIA), along with maximal difference schemes (MDS), have been proposed by researchers, and provide good quality of semi-automatic segmentation. But in these techniques, user intervention is needed in at least 1 of the steps in order to get a proper output. This work proposes a RP-live wire based algorithm which optimizes the segmentation process, and removes the user intervention from the segmentation process in order to produce high quality segmented images with truly automatic segmentation. Our results demonstrate a 20% improvement in overall system speed and 10% improvement in segmentation accuracy when compared with traditional algorithms.

Authors and Affiliations

Pradnya A. Maturkar, M. A. Gaikwad

Keywords

Related Articles

Comparison of Control Methods for Power Quality Improvement Using Dynamic Voltage Restorer and Hybrid Active Power Filter

Power quality is the most common factor which is looked by both power consumers and power suppliers. The power industry has to supply power to the consumers at rated magnitude and frequency and also uninterrupted. But du...

Implementation of a Novel Pwm Dc-Dc Converter

This paper presents the implementation of a soft-switching boost dc-dc converter. Passive softswitching is employed due to its advantages over the active soft-switching and its ability to reduce switching losses. A Lapla...

Design of Wireless Measuring Terminal for Absolute Position and Displacement of Railway Track Based on PSD Position Sensor

With the large-scale application of seamless rail in China, the workload of front-line railway maintenance workers has also greatly increased, and the traditional manual "two times measurement method" has been unable to...

Polycrystalline Photovoltaic Test Set Design

Today, energy demand is increasing day by day. The reserves of fossil fuels are limited and cannot meet this energy which it is needed. To make use of renewable energy resources to provide the energy needs. In this study...

Electricity and Its Dangerous Sides

Dangerous practice of electrical wiring systems and poor products must be thoroughly investigated if electrical fire incidents will be mitigated. Electricity can kill faster than any means and even destroy than tsunami....

Download PDF file
  • EP ID EP406758
  • DOI 10.9790/1676-1305020105.
  • Views 140
  • Downloads 0

How To Cite

Pradnya A. Maturkar, M. A. Gaikwad (2018). Analysis of Random Projection & Live –Wire For Texture Extraction & Segmentation. IOSR Journals (IOSR Journal of Electrical and Electronics Engineering), 13(5), 1-5. https://europub.co.uk/articles/-A-406758