Colour quantisation technique based on image decomposition and its embedded system implementation

摘要: A new colour quantisation (CQ) technique and its corresponding embedded system realisation are introduced. The CQ is based on image split into sub-images the use of Kohonen self-organised neural network classifiers (SONNC). Initially, dominant colours each sub-image extracted through SONNCs then used for entire image. proposed can both components spatial features, achieving better approximation final to characteristics original one. In addition, estimation proper number colours, a algorithm projection first two principal proposed. offers reduction on-chip memory requirements suitable (or system-on-chip) implementation most time-consuming part technique. Applying systematic design methodology developed algorithm, an efficient architecture ARM7 processor high-speed processing less energy consumption, derived.