The fundamentals of underwater photography are light attenuation and colour distortion which are a outcome of light travelling through water. Classic techniques to improve the under-water images are white balance, color cast corrections, dehazing algorithms and contrast stretching based on histogram equalization. Our approach leverages an imperfect, region-based stochastic resonance technique that is tuned by parameter optimization. To change details, the underwater image is first divided into four. There is a unique colour correction process for each zone, followed by exposure to stochastic resonance (SR) technique to amplify weak signals. The parameter tuning is done using a technique to facilitate improvement. Last, a full-image reconstruction is performed by combining the outputs from the color correction and SR algorithms for each region. According to experimental findings, it outperforms previous approaches in visual fidelity and is thus suitable for various underwater imaging applications, including marine biology, underwater archaeology, and ocean exploration.