KEYWORDS: Point clouds, 3D modeling, Mouth, Head, Eye models, Nose, Information visualization, Data modeling, Visualization, Principal component analysis
Haniwa are clay figures and important archaeological materials because they were made during the Kofun period for rituals and talismans against evil. Archaeologists use their knowledge when they observe visual information such as shapes, sizes, ornaments, and noses of Haniwa for classifying who created Haniwa and where they were created. However, classification by observation is largely based on subjective evaluation, and therefore an objective evaluation method is required. In this study, in order to automatically find facial parts of Haniwa, point clouds are projected onto a plane, and facial parts are extracted from the positions of holes representing eyes and mouths. Automatic extraction of facial parts was achieved by changing the threshold value for investigating the size of holes for each model.
Laser scanning and photogrammetry are the methods to obtain the surface point cloud of an object. Laser scanning can measure the actual size, but some missing spots might appear if the laser does not reach the object surface. In contrast, photogrammetry cannot measure the actual size of an object, but missing spots are less likely to appear. By using laser-scanned point clouds and photogrammetric ones together, it will be possible that point clouds of the object’s actual size will be obtained without any missing surface spots. However, it takes time and effort to perform laser scanning and photogrammetry and synthesize the obtained point clouds. Therefore, this paper proposes a method that can automatically measure laser scanning and photogrammetry to improve the efficiency of actual-size point cloud acquisition without missing spots. To confirm the efficiency of our method, the surfaces of an oyster shell and a bird doll are measured as examples. As the result, point clouds without missing spots are obtained.
When a large number of stone tools need to be studied, a unique identification number is assigned to each stone tool to distinguish one from the others. Since most of the current methods of stone tool management are operated manually, identification information may be lost in the process of research. Therefore, a new system is required to provide the identification numbers automatically from actual stone tools. The previous study proposed an identification method that realizes matching using stone tool silhouettes and the ICP algorithm. However, it is difficult to identify stone tools if they are thick because the matching method is based on two-dimensional approach. According to Chida et al., the improvement of performance is achieved using D2 distribution to narrow down candidates as a preprocessing step for stone tool matching. This paper proposes a method that realizes matching including D2 distribution, which enables 3D comparison, with the methods of the previous study to achieve higher identification accuracy.
KEYWORDS: Bone, Point clouds, Image segmentation, 3D modeling, X-ray computed tomography, 3D image processing, Systems modeling, Reverse modeling, Principal component analysis, Overfitting
In recent years, x-rays and computed tomography (CT) are used today to evaluate bone fractures, with the 3D models extracted from the CT images, the dislocation and displacement of bone fragments computed by an assembly simulation using 3D models are necessary for a successful surgical treatment. This paper proposes a method to assemble the point cloud data of the simple fractured bone fragments. Using a multi-layered segmentation method to merge the bone surface, the initial position of the point cloud data is obtained to improve the accuracy of the ICP method.
This paper proposes a quantitative method for calculating and evaluating the facial similarity of human Haniwa based on 2D images generated from 3D point clouds measured from actual Haniwa artifacts. In the field of archaeological research in Japan, it is extremely important to clarify the excavation information, production process, and artistic value of human Haniwa, and to classify (group) each object appropriately. In order to achieve this goal in a quantitative way, Lu et al. proposed a quantitative similarity evaluation method to directly use the 3D measurement point cloud of Haniwa. They reported that the result is to some extent consistent with the clustering result obtained subjectively by archaeologists in the past. However, the computational calculation time required for 3D point matching with repetitive algorithms remains as a problem. Therefore, this paper proposes another simpler method for objective and quantitative similarity evaluation to consume less computational resources by appropriately converting 3D point clouds into 2D images. The paper also shows that the clustering result is as good as the previous research.
Most studies on high-dimensional data preprocessing for data mining and pattern recognition, attention have focused on unsupervised feature selection, which aims to reduce data redundancy and select the most representative features from massive unlabelled data. Previous work on convex nonnegative matrix factorization with adaptive graph constraint has indicated that selecting a few essential features helps replace the original features for clustering. However, the problem exists in how to decide the quantity of these essential features without prior label information. This study proposes a customized evaluation criterion to evaluate each essential feature set. A significant advantages of this method is that the optimal number of features can be determined without knowing their true labels. Experiments were verified by the consistency between the results of our method without label information and that of the comparison method, which required label information.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.