Adaptive Stochastic Resonance Based Low Illumination Image Enhancement Algorithm Based on Symbol Sequence Entropy

Aiming at the low-illumination image enhancement problem, a low-illumination image enhancement algorithm based on symbol sequence entropy and adaptive stochastic resonance is proposed. The algorithm is based on the bistable stochastic resonance model, uses the ant colony algorithm to solve the parameters a and b of the bistable model system, and uses the symbol sequence entropy of the output signal as the measurement standard to judge whether the best stochastic resonance is achieved. The system reaches the optimal stochastic resonance state. In the research process, the iterative threshold of low-illumination images is set, which avoids excessive enhancement of high-brightness areas and ensures that the images have better visual experience. By defining the optimal enhancement quality evaluation index, image enhancement of low-light images is completed. The results show that, compared with the current image enhancement algorithm, the algorithm has better visual information, better maintains the image color, and has good robustness.