Method to Determine Handwriting Stroke Types and Directions for Early Detection of Handwriting Difficulty

Abstract The early detection of children suffering from handwriting difficulty is important for timely intervention to avoid negative impact in learning. In this paper, a method that identified the types of strokes involved in writing Latin alphabets and its drawn direction is proposed to detect handwriting difficulty based on the errors formed during character drawing. The method accepts stroke inputs of point series recorded in xy-coordinates, and categorized the input strokes into three categories, which are simple straight lines, curve lines, and complex straight lines, i.e. combination of two or more simple sstraight lines that are drawn continuously in one stroke. The classification is done based on the angle difference detected between series of points in each stroke. Next, the type of the handwriting stroke and direction to produce each stroke will be identified through the rrelationship of consecutive coordinate pairs. The proposed method works well in classifying and identification of simple straight line strokes, while additional features are needed to improve the accuracy in recognition of curve and complex straight lines.