The Design Of Root Canal Observation System And Image Processing About It

Dental root canal treatment, also known as root canal therapy, is an important means of periodontal disease treatment on clinical dentistry, including a series of steps like cleaning and disinfecting the tooth, the removal of the damaged tooth nerve and bacteria and so on. While taking about the root canal treatment now especially about the people who are still painful after the removal of dental nerve, not only the root canal information inside but also the information of the lowest place of the root canal is very important and needs to be observed and be analyzed. Nowadays, the existing root canal observation system is based on the endoscope and fiber. However, the fiber’s diameter is so huge that the information of the lowest place of the root canal can’t be observed clearly. Moreover, the optical imaging effect is poor resulting in low image resolution and blurred view and it is very difficult for people to understand the meaning of the root canal images.Because there are still some drawbacks in the root canal images in a degree which are acquired from the new observation system for example the resolution is low and the characteristic of the root canal is not obvious and so on and also there is not a commonly used operator can meet the need of root canal image processing in image registration and fusion field, we try our best to reconstruct the original state of the root canal by using image registration, image fusion and three-dimensional reconstruction methods. We proposed new methods basing on scale invariant feature transform and Harris Laplace operator to process the root canal images.In this study, the new methods we proposed are introduced as well as the new observation system. In addition, root canal image registration(including detection and description), fusion and 3D reconstruction of root canal images are completed respectively and the three-dimensional model of root canal is acquired. Meanwhile, the results and the analysis of them in these three main parts are showed also with the evaluations and conclusions. Our work provides a valuable and extensive idea for processing different kinds of images and an explicit direction of processing endoscope images whose characteristic may be the same as that of ours.