A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Gayathri, R.
- Shared Disk Big Data Analytics using Apache Hadoop
Authors
1 Computer Science and Engineering, University of Trichy, IN
2 V.R.S. College of Engineering and Technology, Arasur, Villupuram, IN
Source
Artificial Intelligent Systems and Machine Learning, Vol 7, No 10 (2015), Pagination: 285-289Abstract
Big Data is a term connected to information sets whose size is past the capacity of customary programming advancements to catch, store, oversee and prepare inside a passable slipped by time. The well known supposition around Huge Data examination is that it requires web scale adaptability: over many figure hubs with connected capacity. In this paper, we wrangle on the need of an enormously adaptable disseminated registering stage for Enormous Data examination in customary organizations. For associations which needn't bother with a flat, web request adaptability in their investigation stage, Big Data examination can be based on top of a customary POSIX Group File Systems utilizing a mutual stockpiling model. In this study, we looked at a broadly utilized bunched record framework: (SF-CFS) with Hadoop Distributed File System (HDFS) utilizing mainstream Guide diminish. In our investigations VxCFS couldn't just match the execution of HDFS, yet, additionally beat much of the time. Along these lines, endeavors can satisfy their Big Data examination need with a customary and existing shared stockpiling model without relocating to an alternate stockpiling model in their information focuses. This likewise incorporates different advantages like soundness and vigor, a rich arrangement of elements and similarity with customary examination application.Keywords
BigData, Hadoop, Clustered File Systems, Investigation, Cloud.- Moving Object Detection and Counting Using Fuzzy Color Histogram Features
Authors
1 Department of Computer Science and Engineering, V.R.S College of Engineering and Technology, Arasur, IN
Source
Artificial Intelligent Systems and Machine Learning, Vol 6, No 9 (2014), Pagination: 317-320Abstract
Object detection and counting from video stream is very important for many real-life applications. Existing detection and counting were based on Bayesian regression. We present a efficient object detection and counting based on background subtraction using fuzzy color histogram (FCH), which used effectively for removal of unwanted pixel from the background and capacity of extraordinarily weakening shade varieties created by foundation movements while as of now highlighting moving articles for effective individuals tallying. First, video is converted into frames for processing it to still images to detect objects. Fuzzy C means (FCM) technique used for data grouping, applying along with membership values for clustering with color planes [1]. Foreground and background classified with FCH features by applying threshold value ranges from 0 to 1. Then detected object proceed with morphological process and component analysis for smoothing. Finally, object is counted and we present number of objects in full video.