AI'10 Proceedings of the 23rd Canadian conference on Advances in Artificial Intelligence
| Hi-index | 0.00 |
In this paper we present a syntactic method for handwritten digit recognition using the concept of fuzzy logic. We develop a fuzzy expert system that is capable of handling vagueness in describing each numeral. Our expert system consists of two main parts. First the preprocessing part including the feature extraction step that will provide sufficient information to the inference engine. Second the inference part where the inference engine applies the suitable set of fuzzy rules and aggregates them towards the final decision. By use of our fuzzy expert system, a prior optimization in terms of computing time and memory requirement was performed. An attribution of some intelligence in the decision making as well as an important compression of the amount of information were successively achieved. Some experimental results were provided showing the robustness of our expert in handling confusing samples of handwritten numerals.