A study of using backpropagation and a new neural net algorithm for edge detecting in binary images

<p>Computer image processing often involves three processing stages : 1. detecting and extracting edges from the input image, 2. connecting the detected edges to form objects, 3. using information about the objects to construct and understand a scene. Edge detecting is an important processing stage. The intent of this research is to explore the possibility of using neural computing algorithms for edge detection in binary images. This paper presents the following : 1. the experimental results of using the basic backpropagation algorithm for edge detecting in binary images. 2. the development of a new neural computing algorithm and the results of applying it on edge detecting in binary images. 3. evaluation of the new algorithm. 4. comparison of the new algorithm with a conventional edge detecting method. In order to bring this research under a realistic light, the new algorithm was also experimented on the edge detection of the left ventricle area in a MRI scan of a human heart.</p>

Name thesis:degree_name

Master of Science (MS) in Computer Science

Level thesis:degree_level

Thesis: EWU Only

Discipline thesis:degree_discipline

Computer Science

Year

1992

Author dc:creator

• Tian, Jun

• Other Computer Sciences
• Theory and Algorithms

Statement dc:rights

• Access perpetually restricted to EWU users with an active EWU NetID

Repository record dc:identifier

https://dc.ewu.edu/theses/874

OAI identifier oai:identifier

oai:dc.ewu.edu:theses-1872