改善된 처리 方式에 의한 신호의 표현 = A Study of Processing Method for Signal Display

다국어 초록 (Multilingual Abstract)

A new method is developed to restore the degraded images corrupted by the noise during the transmission, which is called MCSW(Modified contrast-sensitive weights) Filter method. The restored images by this method are compared with images processed by conventional ways, such as Average Filter, Median Filter, and CSW(Contrast-sensitive weights) Filter.
The MCSW Filter divides pixels roughly into two groups by their variance, the one with lower variance and the other with higher variance, and processes each group with a different algorithm, then combines both results together. A computer VAX 11/750 has been used for the simulation.
With the use of the MCSW Filter, the images of higher S/N ratio can be restored successfully, while restoring the images of lower S/N ratio needs the interation of processing more than 3 times. The restored images of lower S/N ratio processed by this mehtod show somewhat larger MSE(Mean square error) but are visually more natural than those processed by the Average Filter method. The simulation results using the MCSW Filter also show that the number of iteration can be reduced to obtain better picture quality in comparing with the other methods.

목차 (Table of Contents)

• Ⅰ. 序論
• Ⅱ. 理論
• 1. 平均値필터(Average Filter)
• 2. 中央値필터(Median Filter)
• 3. CSW필터(Contrast-Sensitive Weights Filter)
• 4. 補完된 CSW필터(Modified Contrast-Sensitive Weights Filter)
• Ⅲ. 結論