Vignetting Correction Based on a Two-Dimensional Gaussian Filter With Harmony for Area Array Sensors

Vignetting is the main factor causing uneven brightness in an image due to the inherent characteristics of the camera sensor. In radiometric correction-based remote sensing sensors, vignetting can lead to uncomparable radiation signals within and between images, and in the processing of remote sensing images, it leads to an unbalanced color when images are mosaiced. Due to distortions of optical devices, uneven response of detectors and other factors, it is difficult to extract the vignetting from complex reference images and correct it accurately. In this study, a low-pass Gaussion filter was selected to extract the vignetting from reference image, and the standard deviation (STD) and mean of the reference images were used as the factor of harmony to evaluate the strength of the filter for attaining optimal vignetting background. In experiments, the method proposed was compared with polynomial fitting method and general Gaussian filter mothed. The results show that Gaussian filter with harmony method achieves the best effect of vignetting correction compared with the other two methods.