We examine the construction factors of color sensitivity information which is perceived from 2 color combinations. To examine the construction factors, we perform the subjective experiment based on the Semantic Differential method which is using 174 pairs of 2 colors and 15 adjective pairs. By applying the factor analysis, 4 kinds of significant factors named "Evaluation", "Activity", "Potency" and "Harmony" are detected. This report also investigates the prediction method to evaluate the factor scores of arbitrary 2 color combination which was not examined. At first, we have carried out the arithmetic mean of factor scores of each color which consists of the 2 color combination. As the result, the performance of the prediction of the arithmetic mean obtains the good estimation. Finally, to improve the performance of the prediction, we employ the multiple regression model which combine the physical quantities of the color stimulus and the factor scores of the monochromatic color.
In this study, we investigate whether visual attributes of objects, such as hue, saturation, brightness, and sharpness of edge, affect on short term memory of visual pattern. In the experiment, the test stimuli composed of four simple object patterns of the same color were presented 800msec and after 10 sec delay, the observer was instructed to recall the test patterns by drawing them on the answer sheet. Achromatic test stimuli presented on the gray background were white, light gray, dark gray, and black. Three different levels of sharpness of edge were prepared for each of them to examine the effect of edge distinctness. Chromatic test stimuli presented on the gray background were red, yellow, green, and blue, and two dull colors of red and blue of which psychometric purities were approximately same as those of yellow and green to examine the effect of saturation. The stimuli of vivid red on the purple and the light-green backgrounds were added in order to find out the effect of the color difference between the object and background. The results indicate that in the achromatic stimuli sharpness of edge of the object does not affect the observer's performance, and in the chromatic stimuli saturation of the object rather than color difference between the object and background is effective for the correct recall of visual pattern.
It has been reported that the constant hue locus of unique red for colored light plotted on the CIE (x, y)-chromaticity diagram is quite different from the constant hue loci of the Munsell 5R and NCS R not only its location but also the direction of the curvature. No study has been done to investigate whether this inconsistency in the red region is due to the difference of color media the difference of stimulus mode, the difference of an experimental procedure, or the individual difference among observers. Thus we measured the hue of color stimuli presented in a surface-color mode (color charts and CRT color stimuli with gray surround) and in a self-luminous mode (CRT color stimuli with dark surround) using two kinds of color naming technique, an opponent type, and a categorical type. The results of six observers showed that the unique red locus in color chart is located near the constant hue loci of the Munsell 5R and NCS R while the location and the curve shape of the unique red locis in the CRT dark-surround is different from it. It is suggested that the mode of appearance is the most effective cause for displacement and distortion of the unique red locus on the chromaticity diagram.
Impressions from natural color pictures have been investigated when their chromaticness were changed. Semantic differential method was used to extract picture impressions and principal component analysis was applied to quantify mutual influence of chromaticness and picture contents on impressions. It is shown that chromaticness influences mainly to the first principal component while picture contents do to the second principal component.
When we look at a picture, we will get some impressions from it. Usually, we do not perceive everything in a picture. We would perceive some parts of the picture depend on our mental conditions or our interests, etc. On the other hand, artists, like painter or photographer, try to express their mental image to the audience by exaggerating or deforming some parts of a picture. Those regions could be attractive for the audience and we could get impressions about the picture from those. In this paper, we consider on a relationship between attractive regions and impressions of a picture.
Kansei Information Processing, which began with imitating human sensibilities, has come to be useful recently in several fields. The authors have so far made various approaches to apply Kansei Information Processing to pearl quality evaluation which needs highly polished sensibility expertise of inspectors. This paper presents several analytical and synthetic methods for Kansei information in developing the pearl quality evaluation system, including psychological evaluation, a factors identification method using sensitivity analysis of layered neural networks, a visual simulator which can form a virtual pearl sample in order to verify and correct the analytical results, and configuration of a psychological scale of essential quality of pearl.
The body and its movement as well as facial expressions send various information and give ideas when we communicate. Communication consists of both a sender and a receiver. The purpose of this study was to investigate what impression a performer (a sender of information) receives through bodily sensations from dance movements. Male and female 6subjects performed Matsumotos'7Motives and they each selected 3words from check list (1) (2). The result was that the students clearly felt the movement pattern and quality of the Dynamic Mv., Sharp Mv., Flowing Mv., Happy Mv. and a certain relationship between each movement and feeling was found out, but the students' bodily sensations were vague in other 3 Mv.
This paper describes a method to analyze kansei information in camera operations. We are using Noh play because it has some convenient features to analyze camera operations. One of them is that it is easy to find out same actions which have different conditions in Noh scenarios. At first, we extracted camera operations of Noh video shots by professional cameramen using 3D camera model. And then, we considered the effects of camera operations by creating and evaluating some new videos which have various camera operations.
Virtual Studio currently uses cromakey method in which a image is captured, and blue portion of that image is replaced by Computer Graphics image or real image. The replaced image must be changed according to the camera motion. This paper proposes a method to extract camera parameters (camera position, rotation and focus) using recognition of pentagonal patterns. At first, we find matching points of two projective images using invariable features of pentagon. Then, we calculate the projective transformation of two projective images and camera parameters using matching points. Simulation results indicate that camera parameters are calculated easily compared with conventional methods.
The timing patterns of a music performance were mesured and analyzed statistically. One purpose of the study was to search for a timing pattern fluctuation. Five artists timing patterns served to search the structure of the piece, by detecting peaks waves in a phrase boundary. We applied power spector analysis to these timing patterns, and we learned the decay of power spector was over minus one. And 20 student listeners were nominated to evaluate performances of the artists by semantic differential method. A principal components analysis suggested that 7 pairs of presentation word were extracted to present the characteristics. Their judgments showed some significant relations to the timing patterns.