My Research

Multi-Object Classification using local features and SVDD


Object recognition and multi-object classification are one of the most difficult problems in computer vision. They have enormous applications in robotic vision, for example robot navigation, localization and mapping. In this research I combined local features with Support Vector Domain Description, a statistical learning method. In my research the improvement has been done for training times. The following graphs give a comparison of the existing techniques and my method for improvement of training times for the classifier. For a full understaning of the research, (Read More)!


Wireless Bomb Disposal Robot (WBDR)

This project was a final year undergraduate project. A robot was developed to aid bomb disposal personnel in handling explosive devices. This was my first robot that I made with four of my team mates. We spent a lot of money on the whole project and at the time of our final presentation the battery discharged!! Talk about bad luck! To have a look at the picutres and the technical report click here...