The purpose of this paper is to characterize the morphological information of craniofacial structure. Firstly, the morphological feature parameters of cephalometric tracings are extracted using image processing techniques. Then, based on these parameters, the morphology of craniofacial structure is classified using K-means method. Afterwards, four morphology related landmarks such as Go, Po, N, and Me for each class are analyze. In this study, fifty tracings are investigated for morphological analysis. From experimental results, the craniofacial structures are classified into six clusters. Each cluster has four images, eight images, ten image, seven image, thirteen images, and eight images, respectively. The range of spatial distributions for the landmark Go, Po, S, N, and Me are all within 4 mm.