ASM technique to detect spines in x-ray images

Abstract

The human spine regularly consists of 24 vertebrae. This project is about to detect the spines by implementing the Active Shape Models (ASM). The approaches ultilized by these technique is manual marking correspondences to make the training image. During this research, OpenCV libraries is the most appropriate material for object detection and image processing. Besides that, this project also conclude that better training shape of images can produce better output. This project used Stasm 4 software library integrated with OpenCV libraries. These configurations need to be test with sample given to make sure that it working properly. Even the training shape cannot detect the spines properly, but it still appears in better condition. In order to make this project success, the research need to improve their effort. Better understanding of OpenCv concept can helps a lot.