Underwater image quality enhancement using fusion of adaptive colour correction and improved contrast enhancement strategy

Underwater images suffer from poor visibility due to low contrast, colour degradation, and noise, making them unsuitable for several visual tasks. While significant progress has been made recently in enhancing underwater images, achieving robust improvements remains a challenge. To address these issues, a new two-step framework combining adaptive colour correction and an improved local and global contrast enhancement algorithm are proposed to improve the quality of underwater images significantly. First, an adaptive colour correction strategy to compensate for degraded colour channels according to compensation factors for restoring the balance in the damaged colour channels is developed. Secondly, an improved contrast enhancement algorithm is presented to enhance the local and global contrast of underwater image and to sharpen the textural details. Finally, these two output images are fused together using principal component analysis (PCA) fusion. Experiments conducted on the established benchmarks for underwater image enhancement, namely the UIEB and EUVP datasets, have demonstrated the capability of the proposed algorithm to produce images of superior quality in terms of qualitative and quantitative performance measures compared to the latest state-of-the-art methods.