Revolutionizing crop phenotyping: Enhanced UAV LiDAR flight parameter optimization for wide-narrow row cultivation

Abstract

This study introduces a method for optimizing flight modes using unmanned aerial vehicles (UAVs) and light detection and ranging (LiDAR) technology, aiming for the efficient and accurate estimation of crop phenotypes in wide-narrow row planting patterns, for cotton. It proposes specialized flight plans that take into account the unique growth stages of cotton and recommends the s-along flight path, which is derived from a detailed analysis of the cross flight path, to facilitate effective and precise data collection. A comprehensive phenotypic index, labeled as ‘P', and a fitting function are developed to describe the relationship between flight parameters, paths, and digital elevation model (DEM) data. The study also introduces two flight strategies, one focusing on accuracy and the other on efficiency, utilizing a sophisticated multi-objective optimization method. Comparative analyses show that the s-along flight path significantly improves efficiency without sacrificing accuracy, compared to traditional cross flight path techniques. The use of high-precision prior DEM data greatly enhances the precision in estimating critical phenotypic parameters such as plant height (PH) and leaf area index (LAI), especially during key stages of canopy growth. By carefully adjusting flight height, speed, and overlap during different growth stages, an ideal balance is achieved between the precision and efficiency of data collection. These strategies markedly increase the accuracy of estimating phenotypic features (P > 0.75) and efficiency (by 42 %–44 %). This research highlights the potential of these approaches in facilitating large-scale phenotypic data collection for precision agriculture, demonstrating their ability to simultaneously improve data quality and operational efficiency. Future research will aim to expand the applicability and robustness of these methods across various planting conditions and crops, further enhancing essential tools for the advancement of precision agriculture practices and development.