부산대학교 도서관

Intra-oral cameras are recommended tools in every dental clinic. The intra-oral camera allows dentists to capture images of difficult-to-reach areas in the mouth. Oral dental applications based on visual data suffer from various challenges such as the environmental effects and the unstable camera movement. We propose an approach, which addresses these challenges, to stitch human teeth images that are captured by an intra-oral camera. First, due to the low density of texture on the tooth surface, normal maps are estimated to reveals the impacted geometric properties of each image inside an area, boundary, and shape. Normal maps are rich with features, which can be used in the stitching process. Second, we investigate the unrestricted camera movement problem. The camera may be moved along the jaw curve with different angles and distances due to handshaking. To overcome this problem, we test each frame, after warping it, and only correct frames are used to generate the panoramic view. The proposed approach shows a similar performance of other cases, in which the camera movement is restricted.