Currently,the panoramic AI image of ancient architecture is affected by real-time changes in lighting conditions.In complex textures and high-contrast areas,there are various fogs with different densities and colors.If the polarization features of images cannot be extracted in time,the dehazing effect of the image will be directly re-duced.To address this issue,a dual-angle polarization defogging simulation for panoramic AI images of ancient archi-tecture was proposed.At first,the original image of ancient architecture was denoised to improve image quality.And then,an atmospheric scattering model was built.Based on the model,the dual-angle polarization image of ancient ar-chitecture was obtained.Meanwhile,the polarization features of the image were extracted.The sky region of the image was segmented.Moreover,the dark channel prior principle was used to calculate the infinite atmospheric light intensity of the image.Combined with the atmospheric scattering model,the defogging enhancement of panoramic AI images of ancient architecture was realized.Experimental results show that the proposed method has good dehazing effect and high performance.
关键词
古建筑全景图像/双角度/偏振去雾算法/图像去噪/大气散射模型
Key words
Panoramic image of ancient architecture/Dual angle/Polarization defogging algorithm/Image denoising/Atmospheric scattering model
维普
万方数据