A new character recognition method using multiple subspace foreach category is proposed. The method is essentially the CLAFIC method.Five subspaces for each category are spanned by vector series expansionsconstructed from five feature vectors extracted from character patternsseparately. The features used are proposed for the pre-classification.The decision procedure of this method is as follows. The feature vectorsare extracted from an input character pattern. The projections of thosefeature vectors on corresponding feature subspaces are computed and fiveprojection lengths are obtained for each category. Five-dimensionalvectors whose elements are the above projection lengths are defined foreach category. The decision rule is to classify the input pattern intothe category on whose five-dimensional vector it has the largestEuclidean norm. The discrimination ability for similar Kanji charactersof this method is examined through a discrimination experiment usinghandprinted Kanji character database ETL-9(B)
展开▼