Reassembly and Restoration of Cracks in Fragmented Images

For the reconstruction of fragmented images and objects a variety of image fragments need to be examined and reassembled. This problem often occurs in several scientific fields, where the manual reassembly of image fragments is very difficult. As the number of image fragments increases and the fragments are irregular in shape, it requires great amount of time and effort for the reassembly. This paper presents an effective reassembly technique for irregular image fragments and the restoration of cracks in the reassembled images. Here the spatial color histogram and the histogram intersection with each fragment is calculated. Then the contour pixel of each fragment is estimated. After the reassembly of image fragments, cracks can be formed. These cracks can be removed using the inpainting technique. The inpainting algorithm, which implements the filling of damaged region with impressive results. The core focus of the reassembly technique is on how to reassemble as many fragments as possible and at the same time to reduce the false matching contour pixels and to remove the cracks formed in the reconstructed image. We have analyzed the performance of the proposed reassembly technique using different number of image fragments and obtained satisfactory results.