Aerial images enhancement using retinex with colour preservation and noise reduction

The aerial image is the process of taking images from the air and it is greatly important in various applications such the knowing the surface of the earth, studying agricultural lands and large pastures such as forests, identifying natural disasters and studying them to avoid them in the future and other areas. However, these images are often hazy. Consequently, the objective of this research was to enhance aerial images by focusing on optimizing the lighting aspect through the application of retinex theory. Subsequently, the enhancement of colors and preservation of vital information were pursued utilizing the Dark Channel Prior (DCP) technique alongside the HSV color space. Finally, a guided filter was employed to effectively eliminate noise from the aerial images. Therefore, analysing the results and comparing the proposed method with several other methods has the best measures of quality for our data reached its values CM(39.577), LOE(159.304) and NIQE(2.721), and it succeeded in improving lighting and retrieving colour information.