【摘要】 暗通道先验去雾算法处理图像易出现局部失真问题,针对这一现象,提出一种单幅户外图像去雾算法。首先,利用四叉树分解的方法得到准确的大气光估计值;然后,结合暗通道图像的直方图分布特征,利用最大化类间方差策略自适应的估计暗通道图像分割阈值,并以此为先验知识优化透射率;最后,使用Gamma矫正提升图像整体对比度。实验表明,该算法能够有效地避免暗通道先验算法在天空区域失真的问题,对比其他算法,恢复图像视觉效果良好,客观评价指标均有所提升。更多还原

【Abstract】 The dark channel prior dehazing algorithm is used to process the outdoor image,the phenomenon of image distortion occurs. In view of this problem,an outdoor image dehazing algorithm is proposed. First,the more precise atmospheric light is obtained by quad-tree decomposition algorithm. Secondly,combining the image dark channel histogram distribution,the strategy of maximum class square is used to adaptively estimate the segmentation threshold of the dark channel image,and the transmittance is optimized by using this as prior knowledge. Finally,Gamma correction algorithm is used to improve contrast of the image. The experimental results show that the method can effectively avoid the image distortion phenomenon in sky area,comparing with other algorithm,both the visual effects and objective evaluation can be obtained the good results by the proposed algorithm.更多还原