Unmanned Aerial System (UAS) Based Cotton Growth Monitoring System

Advances in Unmanned Aerial System (UAS) and sensor development offer great potential for developing scalable, high-throughput phenotyping systems to monitor crops thorughout whole lifecycle. In this study, a novel monitoring system is proposed to track cotton growth over growing seasons using data acquired from UAS. Images were acquired with significant overlap from UAS platform, and 3D point cloud data were generated by applying SfM (Structure from Motion) algorithm to images. CHM (Cotton Height Model) were then generated from the point cloud data. We delineated individual crop grid from an orthomasaic image generated from the SfM process. Maximum height value in each individual crop grid was computed as individual cotton height. These measruements were used to generated growth curve of individual cotton variety by fitting a sigmoid function to the measruements. The experimental result showed that cotton height and growth curve could be successfully estimated using the proposed approach.