An improved Douglas-Peucker algorithm for fast curve approximation

A novel polygonal approximation algorithm for curve representation is proposed in this paper. This algorithm is based on the framework of the Douglas-Peucker algorithm. By integrating the cornerity index as the measure, it iteratively selects the points with local extremum of cornerity index under the control of max perpendicular distance till no points can be marked. With the cornerity index, not only does the corner selection procedure keep simple, the selected corners are more accurate too. The proposed algorithm is extensively tested on various shapes and compared with other methods. The results indicate that it is insensitive to noises. By visual inspection, it can be found that the polygon approximation results demonstrate good shape representation in terms of maintaining essential shape information (corners), which is important in various applications in pattern recognition.