This paper describes properties of Ueda's image enhancement method for line drawings and its merit for practical use. This method can effectively remove discontinuity or mis-connection of lines due to scanning errors. The method is applied the images which are derived from a few parameters (line width and scope of sensor) to evaluate its effect quantitatively and describe that it is more efficient than other methods. As a result, our drawing capturing system with this method can build up drawing databases faster than any other system without it.
Motion-compensated interframe prediction is a key technique for digital video compression method such as MPEG2. In this paper, we propose a method of calculating prediction efficiency using an edge model and impulse model. With the method, the efficiencies of filter type prediction and average type Prediction are calculated and compared. The results are verified by computer simulation using a real video sequence.
In this paper, we propose new computational image sensors which compress image signal in the process of image acquisition. Conditional replenishment is used to reduce the bandwidth necessary for image read-out. We also describe about the design of the experimental chip. This chip has an extensible, parallel, architecture.
This paper describes a new method of representation of volume data with haptic sensation. Recently the volume visualization is often used for analyzing a large number of data. However volume visualized image uses color or intensity of pixel, thus it is hard to display high dimensional volume data. We proposed to give force feed-back to volume model in several way. And then we constructed presenting tool for volume data from computational fluid dynamics research.
sequentially displaying patterns containing specific Key-details, we asked subjects to link and order them by a rule that they must define, and we analyzed their eyemovements. We see first that the dynamic and static repartitions of the fixations follow the cognitive different states all along the encoding. A Total-Local-Global class of eye movements is then defined to characterize the effective visual field and the consecutive scanning initiated to capture some information. The key information position allowed also us to study the parsing habit that gives greater a priori informativeness to the upper part of a stimulus, and perturb the scanning when higher informativeness is placed on the bottom part. scanpaths are then understood as variably perturbed short series of scanning steps that achieve the capture of information following an encoding strategy.
Security of image communication services such as CATV, facsimile, and coming medical image services,etc. will be more important matter, and secure communication should be realized in a simple manner. In this paper, the scrambling algorithm using SFC (Space Filling Curves) and coding algorithm using wavelet transform for still images, which are part of basic technologies to realize secure image communication, are proposed. Also further study items are discussed, analizing simulation results of proposed algorithm.