Volume 6 Issue 10- October 2016


Page Title Full Text
339-343 PET/CT Fusion using Pixel Level Adaptive Weighted Alpha Blending
R. Barani, M. Sumathi

Cancer is one of the primary causes of morbidity and mortality in the developing countries. To infer complex decisions during diagnosis and treatment planning, multi-modality imaging plays an important role. So, accurate anatomic localization of functional abnormalities is desirable. This paper aims to combine the low resolution functional information representing the metabolic activity of tissues from the PET image on the top of detailed, high-resolution anatomical information in the CT image. The paper proposes a new method called adaptive-weighted alpha blending at pixel level for the fusion of PET and CT images. The proposed method is compared with some of the pixel-level spatial domain image fusion algorithms using the non-reference image quality, image fusion and error metrics. It is found that the proposed method excels in performance over the other methods.

Keywords— Adaptive Image Fusion, Non-Reference Metrics, PET/CT Fusion, Medical Image Fusion, Alpha Blending

Full Text PDF
344-348 Novel WebEval in Content Delivery Networks (CDNs) Leakage in Numericable DNS flipped NetProV
T.Devi, O.Madhuri, S.Vandana

A Content Delivery Network (CDN) is a content storage origin customer network, through an optimized and un-optimized connection across various parts of the globe through web, containing copied content in the cached servers, positions cached content to frequently requested customers. A CDN provides multi-client, routing & caching, delivery, security and scalability services to user's global content distribution networks. Most critical issues in multimedia content delivery in internet, public and private networks through content switching engine are transparent breakout, dropping, traffic shaping and  leakage are due to content stream redundant zone, parallel reusable content distributor, high traffic routing clusters and indirect network non-transparent caching. In these issues, the most vulnerable is content leakage. Content Leakage in advanced high-speed wired/wireless web networks is due to the presence of unauthorized user presence in content location, duplicate content delivery and distribution of content replica. Content leakage detection methods evolved since, could not be able to address the solution for parallel reusable content distributors, which provides more content vulnerability in CDN. In this research work, a CDN Domain Name System (DNS) based Web Evaluation intelligence is proposed to resolve the web content leakage through maker agent server (MAS), flipped decision control traffic (FDCT) and  Proxy Caching mechanisms (PCMs) to route multi-cast CDN web content. Our proposed method, limits vulnerable leakages in multiple CDNs through DNS flipped Net Provision Value Consistency (NPVC).
Key Words - CDN, routing & caching, leakage, DNS, Proxy

Full Text PDF
349-355 H-MINE: Scalable Space Preserving Mining
P.Siva , D.Geetha

A simple and novel hyper linked data structure, H-struct, and a new frequent pattern mining algorithm, H-mine, which takes advantage of H-struct data structure and dynamically adjusts links in the mining process. H-mine has high performance, is scalable in all kinds of data, with very limited and predictable space overhead, and outperforms the previously developed algorithms with various settings. First, a major distinction of H-mine from the previously proposed methods is that H-mine re-adjusts the links when mining different “projected” databases and has very small space overhead, even counting temporary working space; whereas candidate generation and test has to generate and test a large number of candidate itemsets, and FP growth has to generate a good number of conditional (projected) databases and FP trees. The structure and space preserving philosophy of H-mine promotes the sharing of the existing structures in mining, reduces the cost of copying a large amount of data and building new data structures on such data, and reduces the cost of updating and checking such data structures as well.

Keywords:  Hyper Structure mining, Frequent Pattern Mining, FP Tree, H-MINE.

Full Text PDF
356-360 Implementation of Rainfall Record and Statistical Application for Android Platform
T. Venkat Narayana Rao, Jella Shruthi

At present every sector in the government is moving towards mobile applications. Rainfall Statistics Android app is to collect day-to-day data about the rainfall conditions in the Telangana state and to cater to the needs of the directorate of economics and statistics as per the national standards of e-governance. This paper focuses on implementation of android app used for data recording, verifying and viewing tasks provided at different level of users (i.e. Mandal, Division and District). Data recording is done by Mandal level user. He can view reports and graphs related to Mandal. Division level user checks attendance and verifies data entered by Mandal user. District user can view rainfall entry and also view reports and graphs. Technologies used to develop this app are Java; Web services (RESTful) with the databases such as SQLite & PostgreSQL and tools used are NetBeans and Android studio.
 Keywords - Android, java, PostgreSQL, SQLite, Web services.

Full Text PDF
361-365 Implementation of Android App for Retail Price Statistics
Dr. Aruna Varanasi , T. Venkat Narayana Rao, , Simhachalam Maithreyi

In the advancing world of technology, mobile applications are a rapidly growing segment of the global mobile market. Mobile applications are evolving at a meteor pace to give users a rich and faster user experience. This paper discusses such mobile application which is used by Directorate of economics and statistics officers in any governmental sector. Retail Price Statistics app is to collect day-to-day data of various retail prices of essential commodities in the Telangana state. To cater to the needs of the directorate and its services as per the national standards of e-governance paper emphasis on data recording, verifying and viewing for different level of users (Mandal, Division and District). Technologies used to develop are Java, Web services (RESTful) and the databases are SQLite & PostgreSQL and with the tools such as NetBeans and Android studio.
Keywords— Android, PostgreSQL, SQLite, Java, RESTful web services.

Full Text PDF
366-369 A Novel Feature Extraction Method Based on Re-Retrieval Phase using Image Processing
T. Venkat Narayana Rao, Akshit Mandala

Artificial intelligence, Machine learning, Computer vision have been an important area of research in the field of Internet of things and Image processing. Image extraction and detection are one of the most important techniques used to identify a particular feature or a pattern in an image. Several algorithms and methodologies were proposed for this purpose using different approaches. Here, a simple algorithm has been proposed for identification of misplaced object or hidden patterns without the use of complex mathematical operations. The system aims at faster search time and identification of the desired object using simple mathematical techniques effectively.

Keywords Digital Image Processing, Machine Learning, Artificial Intelligence, Computer Vision, Virtual Reality.

Full Text PDF