3:30 PM - 5:00 PM
[U05-P06] Ground-based Measurement Methods for Reducing the Cost of Multi-wavelength Imaging by Microsatellites and Drones
Keywords:remote sensing, spectrometer, spectral library, microsatellite
In remote sensing using microsatellites and drones, which are indispensable for next-generation agriculture, multi-wavelength spectral imaging in narrow bandwidths is expected to be used, but at present, it is not widely available. To overcome this problem, it is important to construct a "spectral library," a detailed database of the relationship between light irradiation angles and measurement directions and spectra on the ground in an extremely efficient manner, and to link it with detailed data on crop conditions and yields. However, the creation of such a library requires the generation of a vast amount of spectral data, and due to the time and expense involved, it has not yet been routinely used in actual agricultural practice. We have developed a compact spectrometer to efficiently construct a spectral library. This compact spectrometer can be used not only for hand-held imaging but also for measuring spectra while moving this spectrometer 360 degrees in azimuth and 90 degrees in elevation on a 5-9 m pole to efficiently create a spectral library constructed of a field of about 40 m radius, including angles. In addition, the data can be used to create indexes, check their accuracy, and check for differences in spectra depending on the angle and orientation of the image.
Furthermore, at the stage of utilizing these spectra, if all colors are photographed in the same manner as above, the amount of data and cost will be high, and ultimately the advantages of multi-wavelength spectral imaging cannot be utilized. Therefore, we extracted four color combinations from the spectral library that can discriminate well the state of agricultural crops, developed a camera equipped with filters for those four bands, and devised a measure that can be used routinely in agriculture. This has enabled us to reduce the cost of multi-wavelength imaging to 1/20, and we are now looking at daily use in the field of agriculture.
Furthermore, at the stage of utilizing these spectra, if all colors are photographed in the same manner as above, the amount of data and cost will be high, and ultimately the advantages of multi-wavelength spectral imaging cannot be utilized. Therefore, we extracted four color combinations from the spectral library that can discriminate well the state of agricultural crops, developed a camera equipped with filters for those four bands, and devised a measure that can be used routinely in agriculture. This has enabled us to reduce the cost of multi-wavelength imaging to 1/20, and we are now looking at daily use in the field of agriculture.