Manish Dixit

Improving Mobile Search through Location Based Context and Personalization

In this paper, we present a novel method for deep user profiling using information available on social networks about a users activities and interests. We use this information to personalize the search for the user and make results available via web and low end mobile phones. In addition to using a users profile as a context, we also use the context of location and implement a location based search engine for low end mobile phones. Finally, we combine the two contexts and implement a location based personalized search engine accessible from a low end mobile phone. There has been much work already done in trying to personalize search results based on user profiling. The top search engines attempt to build a user profile based on usage history and then use that to customize search. However, there have been fewer attempts to personalize search based on a deep understanding of a users activities and preferences. In this work, we attempt to do this.

Image Segmentation Methods: A Survey Approach

Image Segmentation is one of basic problem and is an important step in image processing to distinguish important object that is of our interest from unnecessary background substrates. So, quality of this step also affects the subsequent steps of recognition. In this paper, we have compiled Edge detection method, Thresholding method, Region based method, and Clustering.

Power-Aware Consumption Routing in Wireless Sensor Networks

Wireless sensor networks consist of spatially distributed sensor to monitor physical or environmental conditions and cooperatively pass their data to a central location. These kinds of network have come into prominence because they hold the potential to revolutionize many segments of our life from environmental monitoring and conservation to automation in traffic monitoring and health-care business. This paper focus on design, implementation and operation of a sensor network requires the assemblage of various disciplines including embedded systems, hardware, networking, information management and distributed algorithms. Most applications of WSNs require deployment in unattended remote locations. Therefore, power awareness is fundamental to data dissemination and routing in a WSN.

Un-Supervised MRI Segmentation Using Self Organised Maps

The primary aim in brain image segmentation is to perform partition a given brain image into different regions which are homogeneous with some criterion. Magnetic resonance image (MRI) segmentation plays crucial role in accurate representation of white matter (WM), grey matter (GM) and cerebrospinal fluid (CSF) provides a way to identify many brain disorders, such as Alzheimers disease, schizophrenia or dementia. In this paper presents an unsupervised method for MR image segmentation based on Self Organizing Maps (SOMs). The proposed method is consist of five stages these are image acquisition, pre-processing, feature extraction using haralick features, feature selection using principle component analysis (PCA) and tissue classification using SOM. Our proposed method is performed over real MR data provided by Internet Brain Repository (IBSR 2.0) database. Performance evaluation using Tanimoto performance index indicate that the proposed method has good segmentation results. Tanimoto performance index gives mean and standard deviation of 0.68±0.03 for white matter and 0.59±0.05 for gray matter.

Image Contrast Enhancement Using GA and PSO: A Survey

Image enhancement as an optimization process is being used rapidly in Image processing system. To enhance the contrast of the images various common methods are proposed in literature including HE and linear Stretching but they give unnatural contrast and brightness in the images. GA as well as PSO both as an optimization method improves contrast and gives better result. This paper reviews an introduction of several approaches for contrast enhancement in spatial domain as well as in frequency domain using Genetic algorithms and Particle swarm optimization algorithm. The survey of this topic leads to the conclusion that field of the PSO applications is growing fast and improving the contrast using PSO will help to solve many complex problems in image processing in the future.

Route Discovery Protocol for Optimizing the Power Consumption in Wireless Ad-hoc Network

Ad-hoc networks are becoming increasingly popular due to their wide range of uses. Mobile ad-hoc network have Nodes for at no cost to move and arrange them in random order. Ad-hoc networks are suitable for make use of in situation where an infrastructure would not unavailable or to deploy one is not cost effective. In ad-hoc networks MAC protocols are responsible for coordinating access from active nodes. Different MAC protocols with various objectives have been projected for wireless ad hoc networks. Almost of the wired networks based on the symmetric links, which are always unchanging. But this is not a crate with ad-hoc networks as the nodes are movable and continuously varying its position within network. maximize the network life span, is regular purpose of ad hoc network, since ad hoc network nodes are understood to be dead when they are without of battery. Another main objective is increasing the capacity of the network. However we have shown that some schemes degrade the network throughput as well as bring unfairness in the network in order to save energy and to increase the network throughput as compared to the standard protocol. The proposed methods give a very high energy saving and better throughput which increases the battery life of the ad hoc nodes making the system cost effective. Also the proposed schemes bring fairness in the pair wise network throughput.

Enhancing web technology through wiki-shell architecture

The usage of World Wide Web (WWW) and Internet has increased tremendously in the past decade. WWW has changed the way humans now communicate. Initially when WWW was built, it was an application running on the internet. But now web has itself become a platform and numerous applications have been built making use of web-technology stack which run over the WWW. Online shopping portals and online automated systems, for e.g. Amazon and online railway ticket booking, are two hugely popular applications built using the same. Wiki is another such application exemplified by Wikipedia which has become one of the biggest knowledge repositories. Wiki has proven to be an excellent collaboration tool and lot of companies and organisations are using it for various projects and team management. Typically a web application is developed using an application server and a database server like the LAMP stack where Apache is used as web server, MySQL engine is used as the database engine and PHP is used as server-side programming language. In this work, we present a design on how to integrate a Wiki into the current architecture and demonstrate its feasibility. By making use of Wiki engine as a component in the web technology stack, people can build lot of web applications, depending on their needs, in a very easy and convenient way. It can be used in many different scenarios. We have also built a command like application known as Wiki Shell using which people can write queries to fetch information from Wiki Engine.

A Hybrid Facial Feature Optimisation Approach Using Bezier Curve

Image processing is one of the weighty aspects of signal processing where input is an image, which may be a photograph or a video frame from database or any image acquired on real-time basis using frame grabber hardware. The result of image processing is processed image or parameters related to the input image, which is prominently face in case of identity recognition of the human. In the current scenario where security is the principal issue, face recognition is vital importance which can be done by facial features with and without emotions. In the process of face recognition various steps are involved normally as face detection, facial feature extraction and face recognition using any framework. The importance of feature extraction method is foremost in the image recognition techniques. In the proposed approach Harris corner points are considered as significant feature which are optimised using Bezier Curve in turn this low dimension feature vector is used in Image recognition.

Competent Searching for Geographic Information Gathering Using Knowledge Mining

We live in a collaborative world. Arguably, the most productive way of data collection is to recruit a large number of people for data entry. Internet provides a very powerful and versatile medium to bring together such a large number of people With this in mind, we have developed a framework which enables neophyte users to build applications for social co-creation of geographic data, called Geoinformer. Users can create interactive maps with multiple layers to collect geographic information. Geographic information is any data which has a (latitude, longitude) value attached with it. An application can have multiple layers to differentiate features of different types, such as hotels and schools. Geoinformer allows user roles, thereby adding a social abstraction on top of the application, so users can control the process of data collection. We have used the Open Layers JavaScript library to create maps, and the web2py web framework to control the application work flow.