Satellite Image Analysis Using Ai For Deforestation Monitoring In Kalimantan And Sumatera

Abstract

Deforestation in Indonesia, particularly in Kalimantan and Sumatra, remains a significant environmental issue with broad impacts on global ecosystems. This study aims to analyze forest cover change using satellite imagery combined with Artificial Intelligence (AI) technology. The data consist of Landsat 8 and Sentinel-2 imagery from 2015 to 2024. The main methods applied are Convolutional Neural Network (CNN) and Random Forest for classifying forest and non-forest areas. The results show that the model achieved an accuracy rate of 92.4% with a Kappa coefficient of 0.89. Central Kalimantan and South Sumatra recorded the highest deforestation rates, mainly driven by oil palm expansion and mining activities. The integration of satellite imagery and AI has proven effective for early warning systems of deforestation and supports evidence-based conservation policy planning.