一种改进720°全景相机底部图像拼接方法

Improved bottom image mosaic method of 720°panoramic camera

针对支架等遮挡物对720°全景相机底部图像造成遮挡的问题,本文提出了一种基于分步拍摄的消除底部支架遮挡方法。通过对侧面、顶部一次拍摄和对底部两次拍摄消除底部遮挡。在此基础上,针对侧面、顶部、底部图像亮度不一致问题进行了优化。基于全局的匀色算法会导致不同区域的图像互相影响,造成匀色结果的偏差,提出了一种基于区域生长法图像分割的Wallis匀色算法。实验结果表明,本文方法能够消除底部图像遮挡;使用本文算法匀色,图像颜色差异较使用基于全局算法减小59.3%。

Aiming at the problem that brackets occlude the bottom image of a 720°panoramic camera,a method for eliminating the occlusion of the bottom bracket based on step-by-step shooting is proposed.The bottom occlusion is eliminated by shooting the side and top once and shooting twice on the bottom.On this basis,the inconsistency of image brightness are optimized on the side,top,and bottom.The global uniformity algorithm can cause images in different regions to influence each other,causing deviations in the uniformity results.Therefore,a Wallis uniform color algorithm based on region growing method is proposed for image segmentation.The experimental results show that the method in this paper can eliminate the bottom image occlusion.Using the algorithm in this paper to even color,the image color difference is reduced by 59.3%compared with the global algorithm.