一种优化透射率的无人机影像去雾算法研究

Research on a UAV Image Defogging Algorithm with Optimized Transmittance

无人机获取的影像出现不均匀雾,使得传统算法的去雾结果出现细节丢失、色彩失真等问题,很大程度上限制了影像的应用价值。针对该问题,本文提出一种改进透射率的无人机影像去除不均匀雾算法。首先,结合颜色衰减模型和大气散射模型计算透射率,通过改进引导滤波细化透射率;其次,加强图像边缘信息;最后,对去雾后的影像进行增强处理,提高图像的清晰度和对比度。实验结果表明,本文提出的算法对去除不均匀雾取得了较好的效果。计算同幅影像有雾区域和无雾区域的透射率,来去除无人机影像不均匀雾的方法,不仅可以推动去雾算法的理论发展,也可以促进无人机应用于雾天的灾害快速监测和救援。

The inhomogeneous fog in the images obtained by unmanned aerial vehicle(UAV)causes problems such as loss of detail and color distortion in the dehazing results of traditional algorithms,which greatly limits the application value of images.In order to address this problem,this paper proposes an inhomogeneous fog algorithm for UAV image defogging with improved transmittance.Firstly,the transmittance is calculated by combining the color attenuation model and the atmospheric scattering model,and then the transmittance is refined by improving the guided filtering,which in turn enhances the image edge information.Finally,the image after defogging is enhanced to improve the clarity and contrast of the image.The experimental results show that the algorithm proposed in this paper performs well in removing inhomogeneous fog.The method of removing inhomogeneous fog from UAV images by calculating the transmittance of foggy and fog-free regions of the same image not only promotes the theoretical development of de-fogging algorithms,but also facilitates the application of UAVs to the situation of rapid monitoring and rescue of disasters in foggy days.