Object Detection for Autonomous Vehicle

OData support
Nagy Ákos
Department of Automation and Applied Informatics

In today's world, the field of self-driving vehicles is among the most popular topics in the automotive industry. Regarding this area, one of the most challenging tasks for engineers is environment sensing. There are various types of sensors, that can be used for this purpose. Amongst these, the camera can be used most widely. In addition, thanks to the increase in computing power and the fact, that the price of these sensors is much lower now than it used to be, it has become the most popular tool for environment sensing. It can be used to solve problems like lane detection, object detection or traffic sign recognition.

The main goal of my work, is to implement an object detecting algorithm based on motion sensing, via camera. This includes implementating the framework, and creating development support tools, such as logging. In the first part of the paper, I present how the framework is constructed. After that, an overview is given about the theoretical background of object detection, focusing on vision based applications. After that I give a profound explanation about the object detecting algorithm. Finally, I test the algorithm on test videos, examine the malfunctions and explain the reasons for them, and suggest possible solutions.


Please sign in to download the files of this thesis.