Estimation of Curve Similarity Using Turning Functions

The process of classifying objects is a fundamental feature of most human pursuits, and the idea that people classify together those things that people find similar is both intuitive and popular across a wide range of disciplines. Estimation of difference between curves (curve matching) is an useful and often necessary technique in many applications, including: pattern recognition, image object recognition, robotic applications, computational geometry, etc. In this paper, three methods for curve matching using turning functions are presented. While the first two, called plain and polygonal method, are based on a simple adaptation of the existing approaches, the third one, called penalty method, is a new one and tries to overcome some important problems from the first two. The advantages and essential problems of the proposed methods are also discussed. A number of examples are presented to show major differences among the methods and their potential usefulness. AMS Subject Classification: 68T45, 68T10