Peningkatan Kontras Citra Gelap Menggunakan Metode Wavelet Dynamic Stochastic Resonance

Abstract

A technique for enhancing the contrast of dark images using noise caused resonance in the wavelet domain is proposed and investigated in this study. An analogy of a dark image domain with double bistable system parameters is used to model the state of the wavelet coefficients as a double fine particle motion. Repeat scale internal noise due to the lighting sufficiently achieved by tuning approach and detail coefficients using a parameter -dependent equation. Selection of parameters is done by maximizing the signal-to-noise ratio with traditional SR system. Adaptive iteration discontinued on achieving a target value of optimal quality perception, quality contrast and colorfulness. Comparison with a variety of a SR-based techniques and techniques based on non - SR reflects the potential and worth noting the performance of the proposed technique in terms of quality of contrast, color enhancement factor and visual information.