Division of Informatics
Forrest Hill & 80 South Bridge
MSc Thesis #92142 | |
|---|---|
| Title: | Thresholding Images: Neural Net vs. Human Performance |
| Authors: | Pettinaro,GC |
| Date: | 1992 |
| Presented: | |
| Keywords: | |
| Abstract: | In this dissertation the problem of thresholding an image is examined. A threshold is the grey level value used to binarize an image: a pixel with grey level greater than the threshold is considered white (grey level 255), otherwise it is considered black (grey level 0). Thresholding is an important part of some very simple techniques for finding object edges in an image. These are useful techniques because they are cheap and fast, and a robot equipped with a vision system may get a remarkable benefit from it. The literature is full of ways to accomplish thresholding using procedural algorithms. The work developed in this thesis tackles the problem from another point of view: finding a good threshold by means of neural networks. Images are coded and compressed in grey level histograms. A Kohonen Net is used as a comparison with the performance of a human being and different kinds of the histograms are considered and tested with the net. Comparisons with human choice are carried out. |
| Download: | NO ONLINE COPY |
![[Search These Pages]](http://www.dai.ed.ac.uk/icons/dai_search.gif){kind=icon}
![[DAI Home Page]](http://www.dai.ed.ac.uk/icons/dai_home.gif){kind=icon}
![[Comment]](http://www.dai.ed.ac.uk/icons/dai_comment.gif){kind=icon}