Using a boosted tree classifier for text segmentation in hand-annotated documents
Pattern Recognition Letters
Hi-index | 0.00 |
The most common goal of automatic bank cheque treatment systems is the recognition ofhandwritten information. However, in order to do this, it is necessary to use a reliable and efficient process able to identify and to extract the information, which can then be submitted to a further recognition phase. In this paper we present a process for identifying and distinguishing between handwritten information and machine printed text based on a set of local features. This process is based on the characterization of textual elements via properties derived from their content and their shape. The main advantage of this process compared with other similar approaches is that no a priori information of the treated document is used, thus making it more generic and effective.