Estimation of the crop density of small grains using LiDAR sensors

Automatic feedrate control systems have recently been introduced to lighten the job of combine harvester operators by adjusting the driving speed according to the amount of biomass entering the straw elevator. However, knowledge of the crop in front of the machine could considerably improve their performance. Therefore, the feasibility of using two types of Light Detection And Ranging (LiDAR) sensor, and different methods of online data-processing for estimating crop density in front of a combine harvester, has been investigated. Using either of both sensor types to measure the variation in laser penetration depth, good crop density estimations were obtained for different driving speeds and machine vibrations with coefficients of determination (R2) in the range 0.80–0.96. The high-frequency laser scanner gave very good results based on the histogram peak ratio and on a 3-dimensional field reconstruction. Since the amount of incoming biomass is function of both crop stand density and crop volume, a new method for the estimation of crop volume has been proposed which calculates the volume between the ground profile and the crop profile using thin plate spline fitting through the ground level and ear level points. This method was shown to have potential to improve estimates of crop volume.