A LIDAR-based crop height measurement system for Miscanthus giganteus

The objective of this research was to develop a stem height measurement system for Miscanthus giganteus (MxG), to be used as a component in a future Look Ahead Yield Monitor (LAYM). For this purpose, a SICK® LMS 291 LIght Detection And Ranging (LIDAR) unit was evaluated in static and dynamic mode. To eliminate the error caused by inclination angles from undulations in the ground surface and installation, an inclination correction algorithm was developed that improved the measurement accuracy in both static and dynamic mode.In static mode, the sensor was kept stationary and evaluated among various MxG stem densities. The results showed an average error of 5.08% with a maximum error of 8% and a minimum error of 1.8%. The static height measurement approach was also employed to measure the crop height in a 5 × 10 m field, and, compared to manual measurements, an error of 4.2% was achieved.In dynamic mode, the sensor was driven past a field edge to provide a three-dimensional structure of the crop. An Ordinary Least Squares based surface fitting algorithm was applied to generate both the top and ground surfaces of the covered area, resulting in an average crop height. The results showed that the dynamic height measurement achieved an average error of 3.8% with a maximum error of 6.5% and a minimum error of 1.5%.

► This paper describes a system for the real-time measurement of the height of a Miscanthus giganteus crop.
► The system was tested among various crop densities in static and dynamic mode.
► An inclination correction algorithm was developed that improved the accuracy of height measurement.
► Static height measurement achieved an average error of 5.08% with a maximum error of 8% and a minimum error of 1.8%.
► Dynamic height measurement achieved an average error of 3.8% with a maximum error of 6.5% and a minimum error of 1.5%.