Object recognition, a small branch of computer vision, deals with making computers detect and analyze objects in either a picture, a saved video, or a real time video feed. The major challenge in the task of object recognition is obtaining efficiency. There are several object recognition algorithms currently in place which provide astonishing results with varying efficiency. There are several libraries in the world today that provide implemented solutions to help with applications concerning object recognition; two of the most popular of these libraries are OpenCV and TensorFlow. This research focuses on two broad aspects – understanding the working of neural networks, more specifically, convolutional neural networks by experimenting with different models for object classification that use them and implementing real time object detection using pre-trained models and machine learning libraries.


Fox, Nathan


Computer Science; Mathematics


Other Computer Sciences | Other Mathematics


machine learning, computer vision, object recognition, object detection, real-time, neural networks, TensorFlow, OpenCV, convolutional neural networks

Publication Date


Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis



© Copyright 2019 Bhargav Anandkumar Bhalodi