This paper proposes a scheme which combines the conventional technique with a multi-level structure of Thai sentences for detection and segmentation for touching Thai printed characters. The proposed scheme solves problems of both horizontally and vertically touching characters. The complexity of a multi-level structure is employed to classify characters into three zones. The edge detection technique is applied to separate overlapping characters. Then, the horizontal touching characters are determined by using a statistical width of characters. The segmentation point of horizontal touching characters is determined using vertical projection combined with a statistical width of characters. The vertical touching characters are determined by considering the overlapping area of character boundary between zones. The height of line is used to separate the segment of vertical touching characters. Ambiguities are handle by using distinctive features of Thai characters. The effectiveness of the proposed scheme is tested with data from both newspapers and printed documents. The accuracy of 97 and 98 percents are obtained for newspaper and printed documents respectively.

A method for the recognition of handprinted Thai characters input using an image scanner is presented. We use methods of edge detection and boundary contour tracing algorithms to extract loop structures from input characters. The number of loops and their locations are detected and used as information for rough classification. For fine classification, local feature analysis of Thai characters is presented to discriminate an output character from a group of similar characters. In this paper, four parts of the recognition system are presented: Preprocessing, single-character segmentation, loop structure extraction and character identification. Preprocessing consists of pattern binarization, noise reduction and slant normalization based on geometrical transformation for the forward (backward) slanted word. The method of single-character segmentation is applied during the recognition phase. Each character from an input word including the character line level information is subjected to the processes of edge detection, contour tracing and thinning to detect loop structures and to extract topological properties of strokes. The decision trees are constructed based on the obtained information about loops, end points of strokes and some local characteristics of Thai characters. The proposed system is implemented on a personal computer, and a high recognition rate is obtained for 1000 samples of handprinted Thai words from 20 subjects.