Panoramic Image Mosaic Method Based on CBCT Sequence Images

Zha Shan-sha · Zhongguo yixue wulixue zazhi · 2013

Objective: The cone beam computer tomography(CBCT) has the following benefits: high precision, small radiation and rapid imaging. These advantages make it widely used in the dental diagnosis and treatment. If we stitch the sequence of dental CBCT images of teeth and make the teeth of the panoramic image arranged in a line, then the dentists could clearly and intuitively observe every tooth in the panoramic image. The panorama will be very helpful for the dentist's clinical diagnosis and treatment. But how to stitch the CBCT sequence images that we can get the panoranma? To solve this problem, we designed a panoramic image mosaic method for the CBCT sequence images. Methods: In this article, we use a steel ball mould instead of teeth for CBCT scans, then use the CBCT sequence images of the steel ball mould to implement all the experiments.Firstly, we take several frames from the steel ball mould's CBCT sequence images and mosaic them without any registration,then according to the splicing results, determine the range of the overlapped pixels; Secondly, we radomly selected several groups of images, each group has two adjacent frames. For each goup of images, we implement registration experiments of different overlapped pixels, calculate the structural similarity which is based on the weighted information entropy and then measure the registration's results of different overlapped pixels. Comparing the structural similarity, the overlapped pixels whose structual similarity is maximal is the optimal one. Then we calculate the optimal overlapped pixels of every group, and get the mean of all the groups' optimal overlapped pixels; Finally, by using the mean of all the groups' optimal overlapped pixels to registry and fuse the overlapped regions, we mosaic the whole steel ball mould's CBCT sequence images with the method of frame to mosaic, then get the panoramic image of the steel ball mould. Results: In this paper, we use the above method to stitchthe CBCT sequence images of the steel ball mould, then get the panoramic image of the steel ball mould and its neighboring clamping device. The image is clear and the structure is distinct from the background, and these mean that the mosaic method of this article works very well. Conclusions: The experiments show that the proposed method of the panoramic image mosaic which is based on CBCT sequence images in this paper is practical and effective. The panoramic image is clear and the structureis distinct.

Read the paper · More papers on PaperTik