植基於向量量化，一個應用灰色理論的競爭式學習網路於平均值/差值轉換域技術被提出。本篇論文中，灰色理論被應用到一個兩層的修正競爭式學習網路上。其目的在於建立一編碼簿使得介於訓練向量與編碼簿中之編碼向量的灰關聯度最大。影像資訊並經由平均值/差值轉換後，以詳細係數做向量量化。根據實驗結果顯示，基於灰色理論最大關聯準則之競爭式學習網路及於平均值/差值轉換域上所產生的影像壓縮編碼簿具有效性及良好效能。 Based on Vector Quantization (VQ), a Grey-based Competitive Learning Network (GCLN) in the Mean value / Difference value Transform (MDT) domain is proposed. In this paper, the grey theory is applied to a two-layer Modify Competitive Learning Network (MCLN) in order to generate optimal solution for VQ. In accordance with the degree of similarity measures between training vectors and codevectors, the grey relational analysis is used to measure the relationship degree among them. The information transformed by mean value / difference value operation was separated into mean value and detailed coefficients. Then the detailed coefficients are trained using the proposed method to generate a better codebook in VQ. The compression performances using the proposed approach are compared with GCLN and conventional vector quantization LBG method, experimented results show that valid and promising performance can be obtained using the GCLN and proposed approach.
