M. Lazo and A. Sanfeliu (Eds.): CIARP 2005, LNCS 3773, pp. 966 – 976, 2005. © Springer-Verlag Berlin Heidelberg 2005 An Innovative Algorithm for Solving Jigsaw Puzzles Using Geometrical and Color Features M. Makridis, N. Papamarkos 1 , and C. Chamzas 1 Image Processing and Multimedia Laboratory, Department of Electrical & Computer Engineering, Democritus University of Thrace, 67100 Xanthi, Greece papamark@ee.duth.gr Abstract. The proposed technique deals with jigsaw puzzles and takes advantage of both geometrical and color features. It is considered that an image is being divided into pieces. The shape of these pieces is not predefined, yet the background’s color is. The whole method concerns a recurrent algorithm, which initially, finds the most important corner points around the contour of a piece, afterwards performs color segmentation with a Kohonen’s SOFM based technique and finally uses a comparing routine. This routine is based on the corner points found before. It compares a set of angles, the color of the image around the region of the corner points, the color of the contour and finally compares sequences of points by calculating the Euclidean distance of luminance between them. At a final stage the method decides which pieces match. If the result is not satisfying, the algorithm is being repeated with new adaptive modified parameter values as far as the corner points and the color segmentation is concerned. 1 Introduction The aim of this paper is to provide an automatic method for jigsaw puzzle solving. Automatic solution of jigsaw puzzles by shape alone goes back to 1967 [1]. Since then numerous papers have been written, yet few take advantage of color information. The majority of the proposed techniques works on curve matching. Some of them [11] divide the contour of each piece into partial curves through breakpoint. 2-D boundary curves are represented by shape feature strings which are obtained by a polygonal approximation. The matching stage finds the longest common sub-string and is solved by geometric hashing. In this paper we introduce a few new ideas about how color information and shape matching can go along in solving jigsaw puzzles. There are many reasons for someone to work on this subject. Related problems include reconstructing archeological artifacts [2]-[6] and or even fitting a protein with known amino acid sequence to a 3D electron density map [7]. However, what is of most interest is that of simulating the human brain. It is very difficult to create an algorithm as effective as human apprehension yet it is very challenging. In the proposed method, jigsaw puzzle solving algorithm is divided into three main stages. The inputs of the system are images that contain the pieces of the puzzle