Signal Restoration in Random Channels

• Main Authors: Cheng, Kuen-Jenn, 鄭坤振
• Other Authors: Hsien-Sen Hung
• Format: Others
• Language: zh-TW
• Published: 1997
• Online Access: http://ndltd.ncl.edu.tw/handle/11791481811265868519

碩士 === 國立海洋大學 === 電機工程學系 === 85 === In this paper, we consider restoration problem for one- dimension (1-D) signals and two-dimension (2-D) images distored by a random channel. The first-order of the random channel is assumed to be known, but the second- order statistics is unknown. Based on the constrained total least squares (CTLS) originally proposed by Hung, a recursive CTLS algorithm is developed by replacing Rayleigh's quotient with quadratic form as a distance measure, and its restruction performance is also compared. On the other hand, the CTLS algorithm is entended to two- dimensional image restoration problems. From computer simulation results, the CTLS algorithm can restore blurred and noisy images very effctively.