World’s First Fully Automatic Haze Removal and Image Clarification Method for Optical Satellite Imagery

AI_COOL makes satellite imagery clearer, more accurate, and more reliable.

AI_COOL Technology Overview

AI_COOL decomposes the electromagnetic transmission equation into three components — I(X): satellite observation, J(X): true surface radiance, and H(X): haze radiance — providing a clear physical foundation for haze removal and image clarification.

Using multiscale spatial analysis, fractal and other modern mathematical methods, AI_COOL automatically identifies the spatial structure of haze and separates the haze layer from the original image.

With AI-driven, fully automatic quality control, the system restores surface radiance without any parameter input, producing high-precision, haze-free images for quantitative remote sensing, land-cover analysis, and large-area monitoring.

AI_COOL supports rapid, batch-mode processing for optical sensors, including satellite, aerial, and drone platforms, offering a fast and reliable solution for complex and non-uniform haze conditions.

AI_COOL Key Features

Why AI_COOL Is Different

Traditional atmospheric correction software requires many input parameters, expert tuning, and subjective adjustments. AI_COOL eliminates these limitations by directly estimating and removing the haze layer using modern mathematical tools and automated AI quality control — achieving true, high-accuracy, fully automatic haze removal.

AI_COOL Workflow Diagram

AI_COOL vs ATCOR

ATCOR uses physical atmospheric modeling and requires parameter inputs such as visibility, aerosol type, and water vapor.

AI_COOL requires no inputs, uses spatial–fractal analysis to extract haze directly from the image, and includes automatic AI evaluation for clarity optimization.

Example Visual Results

Application Areas