Localization is one of the key problems faced during the implementation of mobile robots. The quantitative measure of the robot position in a Co-ordinate system is useful in a variety of tasks such as delivery, Path following, outdoor as well as indoor navigation. A wide variety of techniques exists in the implementation of Vision based Localization. The key point of Difference amongst these is the technique used to express the Robots Idea of its own surroundings. The Chief Objective of Vision Based localization is to give the Robot an Idea about its location in its currrent Environment.

The commonly prevalent techniques are pixel-based techniques that find the correlation between images or feature-based methods that exploit typical properties of environments such as linear structures. Image retrieval techniques can be regarded as a more general approach to view-registration. The goal of image retrieval is to find images in a given database that look similar to the given query image. Instead of relying on a specific kind of feature like lines or colors, image retrieval systems usually use a combination of different similarity measures that have been proven to yield accurate results in a wide variety of domains.