Rakkan is a Japanese word for the signature of the artist for a completed painting. In this paper, we propose a system that helps the users to identify and retrieve rakkan images of ukiyo-e paintings. The users can select the rakkan region from a digitalized ukiyo-e image via a graphical user interface. The selected rakkan area is binarized using k-means clustering in L*a*b* color space and normalized for retrieval process as a query image. Histogram of oriented gradients and cross-correlation are used for retrieval. The proposed system can be adapted to ukiyo-e authorship identification and ukiyo-e retrieval system for ukiyo-e study and archiving.

This study uses a motion capture system and focuses on up-down dance movements in order to investigate how dancers communicate and interact with each other. In two different experiments, we asked if there were differences in the synchronization of timing, knee height and knee angle, as well as in the correlation of velocities between dance partners (1) when they danced face-to-face versus face-to-back, and (2) when one dancer maintained a constant rhythm as opposed to a variable rhythm. The results of Experiment 1 indicated that timing differences were smaller when a synchronous task was performed face-to-face, and that the cross-correlations of velocities were higher as compared to when the task was performed face-to-back. In both face-to-face and face-to-back conditions, knee height and angle were nearly identical in the two dancers in the synchronous task, whereas the angular variability was found to be smaller in the face-to-face condition. The results of Experiment 2 showed that timing differences were smaller and cross-correlations were higher in the synchronous, variable rhythm condition than in the asynchronous task, whereas knee height and angles did not differ in the synchronous task. This indicated that timing differences decreased when the dancers could see each other, and that they were able to synchronize with each other regardless of changes in rhythm.

This experiment investigated the kansei information and movement characteristics that observers use to perceive expressions of joy in dance. Expressing joy with movements by dancers was recorded. Observers then rated words related to kansei and characteristics of the movements. The results of factor analyses and multiple regression analyses indicated that the expression of joy in dance has particular features. Joy in dance consists of gDynamics, h gExtensity, h gStability, h gFlowing, h gFlexion of the body, h gUpward extension of the limbs, h gForward movement, h gUpward movement, h gDiagonally forward locomotion, h and gWalking. h Stability, movements using space with little downward movement and more upward extension of the limbs, tended to have a greater effect on perceiving movements as expressing joy.

We tried to build a virtual Yamahoko Parade experiencing system as a digital museum research for Kyoto Gion Festival. We collected route data and acceleration data of Hayashibutai, the performance stage of the Yamahoko float, with a GPS logger and acceleration sensors during the parade. We adopted the data to reproduce the vibration using a vibration platform and integrated vibration, sound, and 3D screen to build an experimental system. The analysis of the reproduced acceleration data and the original acceleration data showed that the virtual experiencing system reproduced the vibration of the Hayashibutai. We also received positive feedbacks from the experts participated in our experiment.

This paper describes a method of identifying authorship of Ukiyoe prints by using Rakkan images found in the prints. A weighted direction index histogram method has been used to create the feature vector for Rakkan character analysis. Also the Pseudo Mahalanobis distances were used to judge distances between dictionary templates and test data. The method includes binarization of Rakkan images which has been done by recursively applying the Otsu method, which realized good character segmentation performance, eliminating the influence of stain and smears. We used 100 Ukiyoe prints which were drawn by 10 artists as both templates and test samples. The identification experiment was done using the leave-one-out method. The results of the experiment indicate there is a possibility of successfully using this method for identifying authorship of Ukiyoe.

The purpose of this research is to make a quantitative analysis of Iaido (the Japanese art of using the Japanese sword) proficiency with multivariate data analysis. We carried out experiments of motion capture on Kirioroshi (a straight overhead slash) movement of Roppon-me (a sword thrust using two hands) in Iaido. We can analyze the proficiency of an Iaido practitioner by conducting PCA (Principal Components Analysis) and cluster analysis of parameters of body movement. In addition, it is expected that our research will help Iaido practitioners and masters with Iaido training through giving new information on Iaido movements.

This research was intended to investigate the relationship between body motion and Kansei, or emotional features, conveyed by motion. A work of Japanese traditional dance (Nihon- buyo) in which a performer plays several different roles or characters has been used for a subject of investigation. Through the psychological rating experiment observing a CG character animation in an abstract representation, we found that the observer recognized the impressional factors of the body motions of each individual character or role based on the four aspects: ‘activity’, ‘potency’, ‘gender’ and ‘complexity’. By the analysis of PCA on 22 physical features extracted from body motion data, we found that the variation of the motions could be expressed by four indices; velocity, height of waist, openness of the body and the range of the height of waist. By using a multiple regression analysis, we discovered the relationship between the impressional features given by the motion and the physical motion features.

The purpose of this research is to make a quantitative analysis of leg movement patterns of expert dancer, which cannot be analyzed with a motion capture system alone, using simultaneous measurement of body motion and biophysical information (EMG: electromyogram). We carried out experiments on traditional Japanese dance, Nihonbuyo. We can analyze leg movements of expert Japanese traditional dancer by calculating a center of gravity in human body and co-contraction of muscles in knee and ankle using biomechanics method. In addition, it is expected that our research will help dancers and researchers on dance through giving new information on dance movements which cannot be analyzed with motion capture alone.

This paper describes an application of evolutionary computing to determining appropriate scattering and absorption coefficients in Kubelka-Munk equations for a set of colorants used in paints. The method describes treatment of the solution spaces as a group of environmental habitats in which constant pairs can move from habitat to habitat and survive where they are successful in determining the behavior of the colorants used. This method was applied and tested on a set of 111 paint samples mixed with components including metallic and pearlescent colorants. The method obtained appropriate coefficients for the paint set without any prior determination of the scattering and absorption of the colorants on standard white or black backings.

Content-based image retrieval by specifying composition of images has been studied. Compositional parameters are extracted from images by image segmentation and object extraction procedures. Query is specified through an interactive procedure controlled by a genetic algorithm: First, a user selects one of seven composition templates, each of which represents the typical composition. The system then successively generates patterns of "composition models" according to the templates selected. These composition models describe the pattern of object distribution within an image. A user iteratively selects composition models which properly describe the composition he/she wants. As a result of this interactive and iterative process the system obtains parameter values for an evaluation function. The function are used for judging conformity of images to the query. This method reduces the user's burden of properly specifying the query corresponding to the composition and reflecting user's preference. Experiments proved the effectiveness of the method.

This paper describes a method for indexing and retrieving color paintings, in which information about "principal" colors is used as key for retrieval. Several principal color regions are extracted by evaluating each region obtained by image segmentation in terms of likeliness for forming principal color regions and background color regions. Retrieval of paintings is done by just specifying principal colors used and their spatial relationships, or by specifying "impression word" that indicate the impression of the paintings. Although the method is simple, it must be useful for managing and retrieving pictures in image database system.

Investigates a method of decomposing handprinted visual patterns into several visually significant components. The method is incorporated into an image database system for African ethnological patterns which consist of lines, arcs, solid areas and hatched areas. Patterns are first separated into two sub-patterns, i.e. a solid sub-pattern and an elongated sub-pattern. The elongated sub-pattern is first broken down into small fragments, which are then connected to be integrated into more global features, i.e. line segments and circular arc segments, by bottom-up stepwise strategy. Hatched areas made of parallel lines are detected by using a clustering technique. The method worked well and the result of processing may be used for making description of patterns.>