Research On Color Consistency In Digital Orthophoto Mosaic

During the shooting of aerial images,due to factors such as illumination,time,and physical attributes,phenomena such as uneven brightness and contrast between the interior of the aerial image or the images have occurred,which has caused difficulties in the subsequent series of processing operations.This also directly affects the quality of digital orthophotomap products.This article analyzes and studies the existing method of homogenization of aerial imagery,and improves and innovates it.The main research work is summarized as follows:1)In this paper,the current research status and existing problems of image mosaic technology at home and abroad are introduced in detail,focusing on the problems of mosaic lines and illumination unevenness in image mosaic technology.According to the characteristics of aerial images,we chose the least-squares image matching algorithm.In order to obtain high-precision matching results,we added the collinear conditions and elevation plane physical matching to constrain the algorithm based on the algorithm.The steps were analyzed in detail.2)This article analyzes the vignetting phenomenon caused by the camera hardware and summarizes the source of the vignetting phenomenon and the method of elimination.Finally,based on the grayscale distribution when the image appears vignetting phenomenon,the steepest descent method is added on the basis of the Gaussian surface fitting model,and an improved image vignetting elimination method based on Gaussian surface fitting is proposed,and then the method is applied.Introduced into the image fusion,and finally through experiments show that this method can well eliminate the phenomenon of vignetting.3)This paper proposes a method combining radiation correction and vignetting correction for overlapping images that have been exposed to exposure,white balance,vignetting and nonlinear camera response functions.This method uses Levenberg-Marquardt algorithm to minimize the estimation.Finally,the estimatedparameters are applied to the image.Experiments show that the method can ensure the brightness,contrast and color of the image well,and the calculation time is less,and the complexity is not very high.