A Synthetic Feature Skull Descriptor for 3D Skull Similarity Measurement

3D skull similarity measurement is a challenging and meaningful task in the fields of archaeology, forensic science, and anthropology. However, it is difficult to correctly and directly measure the similarity between 3D skulls which are geometric models with multiple border holes and complex topologies. In this paper, based on the synthetic feature method, we propose a novel 3D skull descriptor, synthetic wave kernel distance distribution (SWKDD) constructed by the laplace–beltrami operator. By defining SWKDD, we obtain a concise global skull representation method and transform the complex 3D skull similarity measurement into a simple 1D vector similarity measurement. First, we give the definition and calculation of SWKDD and analyse its properties. Second, we represent a framework for 3D skull similarity measurement using the SWKDD of 3D skulls and details of the calculation steps involved. Finally, we validate the effectiveness of our proposed method by calculating the similarity measurement of 3D skulls based on the real craniofacial database.

A Deep Learning Method for 3D Object Classification Using the Wave Kernel Signature and A Center Point of the 3D-Triangle Mesh

Computer vision recently has many applications such as smart cars, robot navigation, and computer-aided manufacturing. Object classification, in particular 3D classification, is a major part of computer vision. In this paper, we propose a novel method, wave kernel signature (WKS) and a center point (CP) method, which extracts color and distance features from a 3D model to tackle 3D object classification. The motivation of this idea is from the nature of human vision, which we tend to classify an object based on its color and size. Firstly, we find a center point of the mesh to define distance feature. Secondly, we calculate eigenvalues from the 3D mesh, and WKS values, respectively, to capture color feature. These features will be an input of a 2D convolution neural network (CNN) architecture. We use two large-scale 3D model datasets: ModelNet10 and ModelNet40 to evaluate the proposed method. Our experimental results show more accuracy and efficiency than other methods. The proposed method could apply for actual-world problems like autonomous driving and augmented/virtual reality.