A HANDWRITTEN CHARACTER-RECOGNITION SYSTEM BY EFFICIENT COMBINATION OF MULTIPLE CLASSIFIERS

Handwritten character recognition has been increasing its importance a nd has been expanding its application areas such as office automation, postal service automation, automatic data entry to computers, etc. It is challenging to develop a handwritten character recognition system with high processing speed, high performance, and high portability, be cause there is a trade-off among them. In current technology, it is di fficult to attain high performance and high processing speed al the sa me time with single algorithms, and therefore, we need to find an effi cient way of combination of multiple algorithms. We present an enginee ring solution to this problem. The system is based on multi-stage stra tegy as a whole: The first stage is a simple, fast, and reliable recog nition algorithm with low substitution-error rate, and data of high qu ality are recognized in this stage, whereas sloppily written or degrad ed data are rejected and sent out to the second stage. The second stag e is composed of a sophisticated structural pattern classifier and a p attern matching classifier, and these two complementary algorithms run in parallel (multiple expert approach). We demonstrate the performanc e of the completed system by experiments using real data.