Efficient 2D and 3D image retrieval with enhanced skeleton points

With the latest innovations in computer science, digital images have come to be used in field such as multimedia devices, computer aided design, digital libraries, and electronic commerce. Effective and efficient image recognition has become critically important with the increase in 2D and 3D images. For this reason, studies in fields such as pattern recognition, computer graphics, machine vision, and applied mathematics are being carried out in image recognition. In this thesis, a new shape recognition algorithm proposed by drawing a circle with increasing radii for 2D images, and spheres for 3D shapes is introduced. As every skeleton corresponds to the center of a maximum tangent sphere, the spheres drawn remain partially in the shape after this operation. The ratio between the remaining part of the sphere (circle) in the shape and the total number of pixels, is used to differentiate skeleton similarities and shapes. Also with this study, a comparison has been made with previous techniques and experimental evaluation and operations have been presented to show the effectiveness of the algorithm.