Title :
The properties and training of a neural network based universal window filter (UWF)
Author :
Pugmire, R.H. ; Hodgson, R.M. ; Chaplin, R.I.
Author_Institution :
Dept. of Production Technol., Massey Univ., Palmerston North, New Zealand
Abstract :
Window filters are extensively used in image processing for reducing noise, enhancing contrast, edge detection and other pre-processing operations. This paper describes the development of a general purpose filter. The performance of the filter on simple emulation tasks, and the subsequent development of improved training methods based on an analogy to the task of training people to perform complex tasks
Keywords :
backpropagation; edge detection; filtering theory; image processing; neural nets; backpropagation; complex tasks; contrast enhancement; edge detection; emulation tasks; general purpose filter; image processing; neural network training; noise reduction; people training; pre-processing operations; training methods; universal window filter;
Conference_Titel :
Image Processing and its Applications, 1995., Fifth International Conference on
Conference_Location :
Edinburgh
Print_ISBN :
0-85296-642-3
DOI :
10.1049/cp:19950738
