An Atmospheric Correction Integrated LULC Segmentation Model for High-Resolution Satellite Imagery

TL;DR

This paper presents an integrated atmospheric correction and deep learning-based segmentation model for high-resolution satellite imagery, improving land cover classification accuracy by correcting atmospheric effects in multispectral data.

Contribution

It introduces a novel atmospheric correction method combined with deep learning segmentation, enhancing LULC classification accuracy especially with limited labeled data.

Findings

Effective atmospheric correction improves segmentation stability.

High accuracy in multi-class LULC classification achieved.

Method performs well with sparse labeled data.

Abstract

The integration of fine-scale multispectral imagery with deep learning models has revolutionized land use and land cover (LULC) classification. However, the atmospheric effects present in Top-of-Atmosphere sensor measured Digital Number values must be corrected to retrieve accurate Bottom-of-Atmosphere surface reflectance for reliable analysis. This study employs look-up-table-based radiative transfer simulations to estimate the atmospheric path reflectance and transmittance for atmospherically correcting high-resolution CARTOSAT-3 Multispectral (MX) imagery for several Indian cities. The corrected surface reflectance data were subsequently used in supervised and semi-supervised segmentation models, demonstrating stability in multi-class (buildings, roads, trees and water bodies) LULC segmentation accuracy, particularly in scenarios with sparsely labelled data.