Spectral sensitivities of input devices are important for the colorimetric calivration and evaluation. The problem on the estimating spectral sensitivities of the devices is discussed. A simple method is proposed to get the accurate estimation under the influence of noise.
This paper describes how simple mathematical expressions can be used to generate artificial fractal images. These mathematical expressions are iterated functions which require little information and are resolution independent expressions. The amount of information that such iterated functions require is less than 100 bytes although they can represent complicated artificial fractal images ; for example only four complex numbers are required in the Hata and Hutchinson fractal generation method. We extended the Hata and Hutchinson fractal generation method and Mandelbrot and julia fractal generation methods and derived some extended mathematical expressions for new fractal images.
In CCD camera with on-chip color filter the image data is managed with the aid of auto iris control, auto white balance control and auto focus control functions. These ideas and control algorithm are well devised. However, human beings do these works much flexibly in combination of centered control system in brain with autonomous decentralized architecture in eyes. This inherent control mechanism gives much flexibility to the whole image processing systems. To address this problem and attain real time self-organized control systems, non-leaner oscillator that has good approximation in real nerve action must be exploited.
Avalanche-type image pick-up tube (HARPICON^<(R)>) has been evaluated for digital radiograhy (DR). Theoretical investigation indicates that image signal-to-noise ratio improves to X-ray shot noise limit at the camera noise dominant condition : thick object or low doses. Experimental results show high signal-to-noise ratio and spatial resolution higher than 2.0 lp/mm for 25-cm-thick object at low doses.