Yapay Zeka Hızlı Bakış & Çıkarımlar
AI InsightHence, we conclude that KT system will be a compromising approach in 2D image transmission.
📌 3 Maddede Bu Makale
Abstract
In this paper, in order to compress and enhance 2D images transmitted over wireless channels, a new scheme called Kalman-Turbo (KT) is introduced. In this scheme, the original image is partitioned into 2N quantization levels and each of the N-bit planes is coded by Turbo encoder. After noise corruption of the channel, each of the noisy bit planes is processed iteratively by a joint equalization block, which is composed of low-complexity 2D binary Kalman filter and the Turbo decoder. In our Kalman filter, state equations are modified in order to take the advantage of two-dimensionality, and, previous soft decisions are taken into account to avoid divergence of Kalman filter. Simulation results show that the performance of the KT system is superior at low SNRs when compared with that of plain Turbo decoding algorithms. We can also achieve compression by any selection of N. Hence, we conclude that KT system will be a compromising approach in 2D image transmission.