航空滤光片阵列多光谱图像条带灰度调整算法

Grayscale adjustment algorithm of aerial filter array multi-spectral image

针对航空滤光片阵列多光谱图像中易出现条带灰度不一致问题,提出条带灰度调整算法。利用状态信息对图像进行投影变换,利用SIFT算法提取图像匹配点,计算匹配点坐标差均值作为图像间的平移关系;基于波段图像的重叠区域计算灰度平均值,利用各波段灰度均值比的平均值作为图像灰度调整系数;以中间图像灰度为基准,对图像进行灰度调整,再通过图像裁剪与拼接得到各单波段灰度一致图像。理论分析与实验结果表明:该方法不仅可以有效解决滤光片阵列多光谱图像灰度不一致的问题,而且能够最大限度保持地物光谱信息特征。

Aiming at the problem of inconsistent grayscale of targets in multispectral images of filter array,a new method of the image grayscale adjustment algorithm was proposed.After the projection transformation of image using the state information,the image matching points were extracted by SIFT algorithm and screening,and calculating matching points coordinate difference as a translation relationship between the image;The gradation average was calculated based on the overlap region of band image,and the average value of each wavelength gray mean ratio as the image grayscale adjustment coefficient;Based on the gray level of the middle image,the gray level of the image was adjusted,and then the image was clipped and splice to get the same gray level of each single band image.Theoretical analysis and experimental results show that this method can effectively solve the problem of filter array multi-spectral image grayscale inconsistency,and then effectively improve the application value of images.