基于改进Wallis滤波器算法的无人机影像匀光匀色方法

The method of uniform dodging uniform color of UAV image based on improved Wallis filter algorithm

传统华里士(Wallis)匀光匀色算法通过保证无人机影像内部不同位置、不同影像之间的灰度均值和标准差近似相等来实现影像的匀光匀色处理,但不能解决无人机影像内部存在的亮度不均、反差不一致等现象,影响无人机影像匀光匀色效果。本文提出了一种基于乘性系数和无人机影像像元信息的变换参数改进方法,此方法的改进在尽可能保持待匀色影像色彩损失小的条件下,使得影像不同块之间的均值和方差与模板影像的近似来实现影像间整体色调一致,并且实现了影像内部亮度均值反差一致,取得更好的匀光匀色效果。

The traditional Wallis algorithm realizes the dodging uniform color processing of the image by ensuring that the gray mean and standard deviation of the different positions in the UAV image or those between different images are approximately equal,but it cannot solve the problems existing in the UAV image.Uneven brightness,inconsistent contrast and other phenomena will affect the dodging uniform color effect of UAV images.In this paper,an improved method of transformation parameters based on multiplicative coefficients and UAV image pixel information was proposed.The improvement of this method can make the average value between different blocks of the image as small as possible under the condition of keeping the color loss of the image to be leveled as small as possible.The variance approximation of the source image and the template image was carried out to achieve the same overall tone between the images,and the same brightness average contrast within the image,which achieved a better dodging uniform color effect.