Fall 2013

Race Car Recognition (TAKEN) 

Synopsis: Recognizing race cars by combining different pattern detectors

Race cars have various indicators, such as company logos, race car id number, license plate etc., on them. We can uniquely identify a race car by detecting and recognizing these indicators. In this project you will:

  • Find out what kinds of features might help us distinguish different race cars
  • Investigate what kind of detectors we need to accurately and quickly recognize a race car using one or multiple photographs
  • Evaluate the complexity and accuracy of these detectors
  • Combine these detectors to get a final recognizer

Deliverables: The final deliverables for the project will be a final report on the work done, and implemented algorithms that can perform the proposed methods on the image.

Prerequisites: Knowledge in image processing, programming with MATLAB for signal processing, basic understanding of machine learning and computer vision.

Type of work: 40% research and 60% implementation.

Number of students: 1.

Level: BS, semester project, Computer Science of Communnication Systems.

Supervisor: Gokhan Yildirim ([email protected])