A study on assembling of subpictures with visual servo

In some diagnosis through medical pictures, an object to be observed is much larger than the field of view taken by a camera. This paper presents a method of automatically assembling a set of obtained subpictures into a mosaic picture to reconstruct a complete scene of the object. The method uses the structural information on junctions of branches in the pictures to match a subpicture with each other. In order to speed up the processing, the positions of junctions are computed roughly by fuzzy reasoning within a restricted area of the pictures. Therefore, image processing areas to find the most appropriate relative translation in two overlapped images are taken automatically so as to contain the optimum numbers of junctions. Subimages of the object are captured efficiently through a camera moved by visual servo control. Some experiments, exemplified by assembling microphotograph images of nerve cell networks, are given to show the feasibility of this method