A line-oriented approach to word spotting in handwritten documents

The problem of word spotting in handwritten archives is approached by match ing global shape features. A set of visual templates is used to define the keyword class of interest, and initiate a search for words exhibiting high shape similarity to the model sec. Major problems of segmenting cursive scr ipt into individual words are avoided by applying line-oriented processing to the document pages. The use of profile oriented features facilitates the application of dynamic programming techniques to pattern matching, and all ows us to achieve high levels of recognition performance. Results of experi ments with old Spanish manuscripts show a high recognition rate of the prop osed approach.