基於自適應遮罩大小的暗通道影像除霧方法

Image Haze Removal using Dark Channel Prior Technology with Adaptive Size of Mask

黃韋霖(WEI-LIN HUANG)

指導教授 ： 鄭文昌

朝陽科技大學/資訊學院/資訊工程系/碩士(2019年)

摘要

有霧的影像不僅會降低視覺能見度，也會降低影像處理的能力，因此影像除霧在電腦視覺領域中是相當重要的技術。在這些技術當中較具代表性的方法為He等人提出的暗通道先驗(Dark Channel Prior, DCP)除霧方法，由於在部分情況下DCP方法除霧後容易產生光暈，因此本文為改善DCP除霧後光暈現象提出一個自適應遮罩尺寸的DCP除霧方法(簡稱ADCP)，該方法以有霧影像的梯度反比作為依據並計算出不同的遮罩尺寸，在梯度大的區域使用較小的遮罩尺寸解決光暈現象，而在梯度小的區域則使用較大遮罩尺寸來達到除霧效果，並透過高斯濾波器與函數計算以獲得更好的非線性對應關係，最後使用蟻群最佳化(Ant Colony Optimization, ACO)演算法尋找高斯濾波器與函數的最佳參數，此外本研究也提出一個新的評量指標，並作為ACO演算法的成本評估函數。經實驗結果證實，本研究提出的方法可以有效的改善DCP除霧技術的光暈現象，並且仍然保有良好的除霧性能。

關鍵字

高斯梯度 ； 影像評量指標 ； gamma修正函數 ； 蟻群演算法

並列摘要

Haze images not only reduce visual visibility, but also affect the effectiveness of image processing. Therefore, image dehazing technology is a very important technology in the field of computer vision. Among them, the representative method is the dark channel prior (DCP) dehazing technology propose by He et al. Since DCP is easy to produce halo after dehazing in some cases, this paper proposes an adaptive mask size DCP dehazing method (ADCP) to improve the halo phenomenon after DCP dehazing. This method based on the inverse of the gradient of the haze image and calculates different mask sizes. In the area with a large gradient, a smaller mask size is used to solve the halo phenomenon, while in the area with a small gradient, a larger mask size is used to achieve the effect of haze removal. And through Gaussian filter and gamma function calculation to obtain better nonlinear correspondence. Finally, the ant colony optimization (ACO) algorithm is used to find the optimal parameters of the Gaussian filter and the gamma function. In addition, we also propose a new dehazing performance index and use it as a cost function for the ACO. Experimental results confirm that the method proposed in this paper can effectively improve the halo phenomenon of haze removal image in DCP, while maintaining good haze removal performance.

並列關鍵字

Gaussian gradients ； performance index ； gamma function ； ant colony optimization

參考文獻

[1]E. J. McCartney, “Optics of the Atmosphere: Scattering by Molecules and Particles,” Physics Today, Vol.30(5), 1977, pp. 76-77.

[2]R. T. Tan, “Visibility in Bad Weather from a Single Image,” IEEE Conference on Computer Vision and Pattern Recognition, Anchorage, Alaska, Jun 24-26, 2008, pp. 1-8.

[3]Y. Y. Schechner, S. G. Narasimhan and S. K. Nayar, “Instant dehazing of images using polarization,” Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Kauai, HI, USA Vol.1,Dec 8-14, 2001, pp. I325-I332.

[4]Y. H. Shiau, P. Y. Chen, H. Y. Yang, C. H. Chen and S. S. Wang, “Weighted haze removal method with halo prevention,” Journal of Visual Communication and Image Representation, Vol.25(2), 2014, pp. 445-453.

[5]Q. Zhu, J. Mai and L. Shao, “A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior,” IEEE Transactions on Image Processing, Vol.24(11), 2015, pp. 3522-3533.