CS676 Project

Object Detection and Localisation Using Shape Models - A Supervised Approach



The aim of this project will be to decide if a specific object is present in an image and then find the enclosing boundaries of the object.
This is a supervised algorithm. In the training phase codebook of PAS features will be generated which will be in turn used to generate a model of the object (using an optimisation function) . The minor deformations that may take place in the object will be modeled for the matching phase.
In the test image(of which the features will be extracted first), the possible locations of the object will be decided using hough voting and then in those locations a point matching algorithm will be used to map the model for that object to the features of the image (using another optimisation function).
IM



Link to the reference paper
List of references(bibTex file)

Paper review
1.HTML
2.PDF

Final Report

Source Code