Skip to Main Content
Artificial immune systems[1-2] are highly distributed systems based on the principles of the natural system. In this paper, handwritten Icelandic character recognition strategy using artificial immune system was proposed and carefully experimented. With 73 feature coefficients extracted from 24*24 handwritten Russian uppercase character image using 36 sub-meshing coefficients, 24 traversing-times coefficients, 1 segmentation-times coefficients and 12 vertical projection coefficients as its feature vector, 32 antibody libraries for 32 character category were trained and built to recognize handwritten Icelandic characters with artificial immune algorithm. The contrast experiment was done using BP neural network. The experimental results indicated that the artificial immune system model has more advantages than BP neural network in handwritten Icelandic uppercase character recognition.
Information Technology and Artificial Intelligence Conference (ITAIC), 2011 6th IEEE Joint International (Volume:2 )
Date of Conference: 20-22 Aug. 2011