Tone Reproduction Algorithm Based on Local Contrast Preservation

In order to address suboptimal contrast and attenuated detail which are introduced by existing tone reproduction algorithms, a tone reproduction algorithm is proposed based on local contrast preservation. Firstly, the proposed algorithm takes the second-order norm as the clustering criterion. Dynamic programming is employed to find the optimal solutions for global tone reproduction. Secondly, linear equations are deduced by equivalent local contrast that is estimated by local edge-preserving filtering. The solution of the equations is the reproduction image that preserves local details. Comparison experiment results on TMQI database indicate that the mean opinion score, tone-mapped image quality index, and mean local weber contrast of tone mapped image reproduced by the proposed algorithm achieve 0.898, 0.896, and 1.645. Thus, the proposed algorithm ensures that the overall contrast of the reproduced image and the reproduced image is full of rich detail on all parts of the picture.