Year : 2017
Mentors : Tanish Zaveri
Students involved : Jonti Talukdar (14BEC057), Bhavana Mehta (14BEC028)
Project : Link
Abstract: This report presents a conclusive treatment of the algorithmic approach used in Image Processing with the specific context of their implementation of HAR in Python, a high level language. Image Processing is a vibrant and diverse field with immediate applications in a wide variety of branches. Python being an easy and user friendly language provides a solid framework for the development of complex algorithms for processing images and a robust platform for processing large amount of data. This allows the developer to move from the mundane task of focusing on syntactical problems to the highly rewarding and important task of developing new models to integrate algorithms, develop systems and process data. The rise of Machine Learning has led to the growth of using such algorithms to automate image processing tasks which include automated recognition, detection and classification of images. This has also led to the growth of research and development of many new models and techniques in computer vision, which find applications in many spheres of life. The problems of image filtering, feature extraction, classification are discussed as well as the challenges observed are stated. With the growth already complete in the field of image processing, new trends still emerge and provide scope for potential development in the field of mathematical modelling for complex learning systems based on images. These include Human Action Recognition, Optical Character Recognition, Pattern Recognition and many more. Optimization in the field of Machine Learning can be done by focusing on developing new techniques to extract complex features easily and to classify large amount of data in a small number of instructions.