AE—Automation and Emerging Technologies: Novel Imaging Spectroscopy for Grass Sward Characterization

Attempts to improve grassland management may benefit from the use of new sensing techniques, such as imaging spectroscopy. In order to explore the potential of hyperspectral imaging spectroscopy for rapid and objective characterization of grass swards an experimental prototype has been developed. From a height of 1 m, the system detects reflections at high spatial (0·16–1·45 mm2) and high spectral resolutions (5–17 nm) at wavelengths between 405 and 1659 nm. Detailed information is provided on design, characteristics and test results. Results show that reflection intensity is related to height position in the sward and leaf angle. Grass leaves were recognized accurately. Images of a sports field and a production sward with similar ground cover fractions could be easily distinguished. With the described system, canopy structure can be characterized by ground cover, the distribution of reflection intensity and image texture. Spectral characteristics can be obtained at leaf level.