针对雾天模糊图像对比度低的特点和数学形态学增强后的远景效果不佳的缺点,提出了基于模糊对比度和数学形态学相结合的雾天图像清晰化算法。该算法采用数学形态学中的高低帽变换对模糊图像进行预增强处理,将处理后的图像从空域映射到模糊特征平面,然后利用新定义的模糊域隶属度函数进行模糊对比度增强变换。实验结果表明,采用本算法处理后远近景对比度明显提高,视觉效果也明显改善,图像变得更加清晰。
There are some shortcomings for the fog fuzzy image, such as the low contrast, blurred, and poor effects after the mathematical morphological enhancement. A new image clearness algorithm based on the fuzzy contrast and mathematical morphology is described. Using the previous fuzzy contrast enhancement algorithm, a new fuzzy membership function is defined, the image is changed from spatial domain into fuzzy field directly through it; therefore a fuzzy feature plane is obtained. In the fuzzy feature plane, the image could be enhanced. Lately, the image is taken back to its spatial domain. Then it is combined with the top-hat transform and bottom-hat transform of mathematical morphology pre-treatment. Finally, the enhanced image is obtained. An experiment for the fog image is presented. The results show that the algorithm has higher contrast and better visual effects. It is quite obvious that the algorithm is able to raise the contrast at the border and background and improve visual effects. In the other hand, this algorithm is good at making fuzzy image details become clearer, the noise is also suppressed, and there is no over enhancement, particularly in the sky area. Therefore, the proposed algorithm is an effective algorithm to clear fog image.