Abstract
In this paper, we present the implementation of a Haar-Cascade based classifier for car detection. This constitutes the detection sub-module of the framework for a Smart Parking System (SParkSys). In high density cities, finding available parking can be time consuming and results in traffic congestions as drivers cruise to find parking. The computer vision based smart parking solution presented in this paper has the advantage of being the least intrusive of most car sensing technologies. It is scalable for use with large parking facilities. Functional code from OpenCV was used in conjunction with custom Python code to implement the algorithm. Our tests show mixed results, with excellent true positive detections along with some with false negatives. Remarkable is that the classification algorithm learnt features that are common across a wide range of objects of interest.