针对传统直方图均衡算法存在过增强、图像细节表现不稳定等现象, 提出了一种基于变异系数反馈调整控制的自适应双平台直方图均衡化算法。该算法在反馈控制原理的基础上, 引入了一个直方图归一化变异系数作为反馈参数, 用来不断修正平台阈值与期望值之间的偏差, 从而自适应地获得最优双平台阈值。定性定量实验结果表明, 本文所提出的红外图像均衡化算法在不同场景的增强效果中表现良好且稳定。

The traditional histogram equalization algorithm is over-enhanced and the image detail is unstable. Therefore,an adaptive dual-plateaus histogram equalization algorithm is proposed, based on the variation coefficient combined with the feedback control principle. On the basis of the feedback control principle, our proposed algorithm introduces a histogram normalized coefficient of variation as a feedback parameter, to continuously correct the deviation between the plateaus threshold and the expected value to adaptively obtain the optimal dual-plateaus threshold. The results of the experiment show that the contrast enhancement algorithm proposed in this paper performs well and is stable under different scenes.