Hierarchical approach for object detection using shape descriptors

Automatic object recognition is a fundamental problem in the fields of computer vision and machine learning, that has received a lot of research attention lately. While there are different methods, that build upon various low level features to construct object models, this work explores and implements the use of closed-contours as formidable object features. A hierarchical technique is employed to extract the contours, exploiting the inherent spatial relationships between the parent and child contours of an object. Fourier Descriptors are used to effectively and invariantly describe the extracted contours. A simple hierarchical, shape label and spatial descriptor matching method is implemented, to determine the nearest object-model. Multi-threaded architecture and GPU efficient image-processing functions are adopted making the technique efficient for use in real world applications. The technique is successfully tested on common traffic signs in real world images, with overall good performance and robustness being obtained as an end result.