This is the project for Digital Image Processing course
Advisor: Prof. A.V. Subramanyam
A system that takes an image as input, checks whether the image contains a traffic sign or not, if it does, then outputs the meaning of the traffic sign. The problem of Real-time traffic sign recognition can be broken into two parts – the first part, Sign Detection, checks whether an image contains a traffic sign or not. The second part, Sign Classification, comes into the picture if the result of Sign Detection is positive and gives the meaning of the detected traffic sign. We experimented with machine learning techniques for Sign Classification, but the data required to do so is limited to only some traffic signs. We also applied another approach called Content-Based Image Retrieval which matches real-time traffic signs with virtual images of traffic signs using distance calculation between different types of features of an image.
Team: Shreya Sharma and Sarthika Dhawan