A Color Correction Method Based on Incremental Multilevel Iterative Histogram Matching

The stitched images produced by multicamera systems often suffer from unnatural seams and poor color consistency, which seriously impact the accuracy of post-processing and degrade the viewing experience. The existing color correction solutions still have encountered issues such as inadequate correction accuracy and color loss. To solve these issues, we have proposed an incremental multilevel iterative histogram matching (IMIHM) color correction method. The proposed method effectively addresses color deviation by partitioning the image using an adjustable sliding window and performing incremental iterative matching at each level. Additionally, a weighted fusion algorithm is employed to ensure the generation of seamless and color-consistent images. This method has the advantages of both color consistency of global color correction and high accuracy of local color correction. Our proposed method has been successfully applied to a
$5\times 5$
camera array system, and experimental results demonstrate that it can eliminate color deviation significantly. Compared to the state-of-the-art hybrid histogram matching (HHM) algorithm, our method exhibits superior performances, with approximately 61.6% improvement in histogram similarity (HSM), 11.6% improvement in mean-squared error (MSE), and 4.2% improvement in peak signal-to-noise ratio (PSNR).