| 摘要 |
<p>PURPOSE:To reduce the quantity of data by multiplying each threshold value of a quantized matrix set previously by a correction coefficient including a prescribed coefficient when an image is compressed. CONSTITUTION:A correction coefficient is defined as 'aX(i+j)', for example, and the increase/decrease of the threshold value can be largely controlled for the correction coefficient as the frequency component is increased. That is, a coefficient (a) is decreased for the reduction of the quantizing width when a large high frequency component is included in an image to be compressed. Meanwhile the coefficient (a) is increased for the incremeat of the quantizing width when the image to be compressed includes a small high frequency component. Thus, the quantization is attained with a quantizing matrix different from the one that is previously set in accordance with the contents of the image to be compressed. Furthermore it is required to transmit or record only the coefficient (a) in the correction coefficient 'aX(i+j)' used for correction of each threshold value of the quantizing matrix.</p> |
