Development and Application of Multispectral Drone Image Pre-Processing Module for Agricultural Monitoring

In order to address issues related to agricultural productivity and sustainability caused by climate change, the importance of agricultural observation monitoring is increasing.Remote sensing data at the field, county, and national levels are being collected using various remote sensing platforms.In particular, drone imagery is becoming an attractive platform with high resolution compared to existing satellite and aerial imagery, allowing for observation of desired regions at specific times, thus enabling immediate responses to agricultural environmental changes.Pre-processing for crop condition analysis is complex and time-consuming, limiting the timely production of information.This study developed and evaluated an automated pre-processing module for multispectral drone imagery in agricultural observation systems to reduce user dependency and enhance standardization and scalability.The pre-processing module consists of a reflectance normalization module for quantitative reflectance extraction and a high-speed geometric correction module for producing mosaic images in near real-time.Through the development and automation of pre-processing modules required for multispectral drone imagery, the efficiency and accuracy of data processing were improved.It is expected that the use of these systems by non-expert farmers will increase, supporting real-time agricultural observation, boosting reliability, and enabling quick agricultural decision-making.