Reconstruction the complete image from random scattered pieces of image

Abstract

Combining all image pieces which are randomly scattered to reconstruct the initial (original) image takes so much time. For instance, combining the all image pieces used in tiling is a tedious and time consuming task. For this purpose, the hardness of the task is able to be decreased and the task duration is able to be shortened using the developed image processing optimization algorithms. In this study, the reconstruction of initial image is implemented using image processing methods from randomly scattered image pieces. In the first stage, the edge improvement is completed to reach the initial image position from randomly scattered image pieces. Then, the rotating angles of image pieces are determined with the support of corner finding algorithm. The initial image positions of pieces are determined and positioned to their original places by individually comparing the prescribed image pieces with the initial image pieces.