Date of Graduation

12-2023

Document Type

Thesis

Degree Name

Bachelor of Science in Computer Science

Degree Level

Undergraduate

Department

Computer Science and Computer Engineering

Advisor/Mentor

Gauch, John

Committee Member/Reader

Panda, Brajendra

Committee Member/Second Reader

Le, Ngan T. H.

Abstract

OpenCV, and Computer Vision in general, has been a Computer Science topic that has interested me for a long time while completing my Bachelor’s degree at the University of Arkansas. As a result of this, I ended up choosing to utilize OpenCV in order to complete the task of detecting road-lines and mapping roads when given a wide variety of images. The purpose of my Honors research and this thesis is to detail the process of creating an algorithm to detect the road-lines such that the results are effective and instantaneous, as well as detail how Computer Vision can be applied to abstract ideas of detecting objects near the same level that a human would.

The detection and mapping of road-lines was executed via an algorithm that incorporated several different parameters and specifications, all of which went through many iterations before being somewhat optimized to the state it is currently in. Furthermore, the thresholding algorithm that I developed would allow and adapt for inputting many different images, as well as active video functionality. This algorithm manually reads through every pixel in the image/frames of a video and detects row patterns by preforming a multitude of custom mathematical calculations on the BGR color values of the image/frames, resulting in the filling of the bounds of the lines by row and thus detailing where the road is.

Keywords

Thresholding; Computer Vision; OpenCV pixel operations; Video Processing

Download

Share

COinS