LAUSR

The objective of this research was to eliminate the influence of environmental stray light on the measured spectral reflectance in a spectral sensor system used on a weeding teleoperated robot for rape fields. The system mainly consists of a light source, signal acquisition, and data processing. In this study, optical modulation technology and discrete Fourier transform method were used to carry out experiments based on four characteristic wavelengths. The results show that the experimental system can obtain a stable reflectivity value no matter whether the stray light changes slowly or dramatically in the field environment. Furthermore, the calibration equation of spectral reflectance was obtained by curve fitting based on a FieldSpec3 portable spectrometer. The determination coefficients are all close to 1, and the root-mean-square errors are small. Through verification experiments, the results show that the average measurement error is less than 5%. Hence, this proposed method would be very helpful in improving the accuracy and efficiency of the spectral reflectance measurement, which in turn provides a theoretical basis for weed identification during in-field weeding by teleoperated robots.