Comparative Study of Region Localization Methods with Image Enhancement for Computer Vision

Authors

Presenter(s)

Quinn Graehling

Comments

This project reflects research conducted as part of a course project designed to give students experience in the research process. Course: ECE 595 48

Files

Description

Region localization is one of the main tasks within computer vision and pattern recognition. Early forms of region localization relied on basic pixel intensity thresholding while later versions used machine learning methods to locate and segment objects of interest within an image. Today the region localization fields are dominated by adaptive progressive thresholding methods, region growing segmentation and neural networks designed for semantic segmentation. With the creation of new image enhancement methods, such as the Retinex method, and with the increase in demand for quick image segmentation for use in artificial autonomy, the need for methods that can quickly and accurately segment images has grown exponentially. This presentation aims to analyze modern image segmentation methods and determine which method performs the quickest and with the highest accuracy. This presentation will also look at the difference in results between segmentation of raw images and segmentation of images with contrast enhancement via Retinex image enhancement.

Publication Date

4-22-2020

Project Designation

Course Project

Primary Advisor

Vijayan K. Asari

Primary Advisor's Department

Electrical and Computer Engineering

Keywords

Stander Symposium project, School of Engineering

United Nations Sustainable Development Goals

Industry, Innovation, and Infrastructure

Recommended Citation

"Comparative Study of Region Localization Methods with Image Enhancement for Computer Vision" (2020). Stander Symposium Projects. 1962.
https://ecommons.udayton.edu/stander_posters/1962