Developable surfaces are important for representing and understanding geometric features of 3D models. Some methods are required that surface pattern of a relic should be easily observed in archaeology area. This paper introduces a system to visualize the development of relic's surface pattern based on points, which does not require reconstruction of a mesh model and relies only on the information of points to develop the relic's surface. After a point cloud of a relic is segmented to separate sections, points of each section are projected to a developable surface, then the points of relic's surface are unfolded into an image plane. In our method, the number of segmentations can be controlled by an archaeologist in an interactive interface. The more segmentations a point cloud is divided into, the higher precision the image plane is shown in adjacent area. To observe the surface pattern easily, a developed plane can be interactively rotated on the prime meridian, which is extracted from a relic's surface. This approach can help archaeologists move a specific area of interest to the center of the relics.

View full abstract

Relics such as stone implements and earthenware are analyzed and researched archeologically according to point cloud data obtained by three-dimensional measurements of the relics. Relics are archeologically precious and some relics measured in this study are accompanied by a risk of damage caused by movement. In addition, a large quantity of relics are excavated from archaeological ruins. As such, it has a high cost to move the relics for measurement. Therefore it is effective to measure excavated relics locally and to transfer the measured data via an internet. However, as the measurement precision gets better it will result in a greater number of points which will increase the point cloud data size. Also, the long transmission times for these large-scale point cloud data becomes a problem. One of the methods to solve this problem is to use data compression to express point clouds with smaller amount of data. This study examines the method to compress point cloud data by approximating flake surfaces in stone implements with trimmed surfaces, which can be expressed with the smaller number of control points than the total number of points in a point cloud. In addition, this study examines the method to restore a point cloud data from the trimmed surfaces. We applied this method to a point cloud data measuring in a 0.1mm pitch and a good result was obtained.

View full abstract

In this paper, we propose a procedural animation method of diffusion phenomenon of sand ascending in water. In first step, we create velocity field of sand ascending in water using the FLIP method. And it is a three types of "for coarse sand," "for the medium sand", "for the fine sand," based on the classification method of Geotechnical Society criteria. Next, we combine the noise based simulation to represent a unique movement by interfering with the sand movements and various sand and water are affected by changes in the complex water flow. In our method, we can add the details of motion by adding particles individually to a noise in order to reproduce the like interference with water. And we can reduce the calculation cost of the density by using the particle volume rendering. Our simulation method, it is possible to efficiently visualize the sand ascending in water.

View full abstract

A method for converting digital illustrations to stereoscopic images is proposed. We noticed that digital illustrations are drawn using multiple layers. The popular method to make stereoscopic images is a technique using 3DCG. In this case, creation of a 3D model of a scene is required. Technique using a depth map is another popular method to make stereoscopic images. The goal in this paper is to easily create stereoscopic images using only the techniques of drawing general 2D illustrations. In the proposed method, the layers are classified into four type of elements: "floor", "wall", "object", and "other". Stereoscopic images are created by transforming these elements geometrically.

View full abstract

In cartoon rendering of 3DCG, methods for generating a contour lines are classified into the method using the image processing or the method using the three-dimensional shape of the models. By selecting G-Buffers, the method using the image processing can accurately detect intended lines of users, but the lines are mechanical impression by the uniform thickness. The methods using the three-dimensional shape of the models can adjust thickness of the lines, but detection of the lines is dependent on the rendering algorithm and it can't detect the intended lines of users. In this paper, we separated contour lines process into detection and rendering. Our approach uses image processing for detection and model shapes for rendering. It can detecting intended line of users and draw exaggerated lines by changing the thickness of lines. In addition, we implemented a graphics API pipeline to run our method in real-time, and also examined a technique to accelerate processing speed.

View full abstract

A user controlled texture synthesis technique was proposed to allow a user to control texture synthesis. The technique creates an output texture (output image) from an input texture (input image) by using a target image painted by a user. When an object and its background exist in the input image, the user controlled texture synthesis technique usually treats the pixels of the object and those of the background equally without treating the contour of the object intentionally. Therefore, if we want to synthesize only the input pixels of the object onto a different background, it is difficult to reconstruct the characteristic contour of the object on an output image by giving a simple contour roughly painted on a target image. In this paper, we propose a texture synthesis technique to reconstruct a characteristic object contour intentionally by expanding the boundary of a painted region on a target image suitably and selecting optimal input pixels to synthesize by the evaluation using boundary pixels and boundary distances.

View full abstract

A visual expression called Energy-Wave, which shows light emission from high-energy fields, is popularly used in creative contents such as animations and computer games. This research proposes a new method for representing energy-Wave in real-time 3D graphics. Conventional energy-wave rendering techniques use the animation texture or the particle animation. However, the animation-texture method restricts the view point, and the boundaries of the energy-wave appear too clearly. The particle-animation method is not suitable for showing a dense energy field and not appropriate for covering a wide region. This paper presents a new method in which the energy distribution is described as a continuous scalar function. The new method is advantageous over the conventional methods since there is no restriction of the view point, resolution, and the covering region. The proposed method also utilizes GPGPU, and the output images are dynamically created from small input and can quickly deform the shape of the energy wave. The previous work of this research was able to deal with point-centered and line-centered energy-distribution functions. The new method can deal with torus-shaped and quadratic curve-centered distribution functions quickly.

View full abstract