Conference paper
Neural network architectures for rotated character recognition
Hiroyasu Takahashi
ICPR 1992
Using simple ideas from techniques used in the numerical solution of partial differential equations, the concept of implicit convolution is introduced and analysed. Its properties on one-dimensional finite-length signals are first considered. Then, by appealing to the theory of methods of split-operator type, the use of implicit convolution for image enhancement is presented. This product-form approach can easily be extended to images in higher dimensions. © 1985.
Hiroyasu Takahashi
ICPR 1992
George Saon
SLT 2014
Achint O. Thomas, Nalini K. Ratha, et al.
ICPR 2000
Tanveer Syeda-Mahmood, Dragutin Petkovic
Signal Processing: Image Communication