huanleo/Object-Detection

Introduction The project presents an efficient method for object detection and demonstrates its use for real-time vision on-board vehicles. The work is closely based on Paul Viola’s face detection method [1], except that the system here was trained to detect objects related to traffic in order to assist the auto pilot on smart vehicles. The idea was adapted from D M Gavrila’s paper [2] REAL-TIME OBJECT DETECTION FOR "SMART" VEHICLES. Objective The target is to create Haar feature-based cascade classifiers by training the system with an appropriate amount of traffic related information in order to detect vehicles, pedestrians and traffic signs in real-time. Thus, this system in addition to driving assistance/collision avoidance, could further find its application in tracking and recognition.

Python