| 研究実績の概要 |
We developed a fluorescence image acquisition system for analyzing the phenotype of green peppers. The fruits had some fluorescence substance whose excitation wavelength was 365 nm and emitted blue fluorescence, which can be captured by cameras within the visible light spectrum. Notably, the "Manganji sweet peppers" in Kyoto, Japan, represent a distinctive variety with significant demand for agricultural big data under their regional brand. In response to the need for individual green pepper traceability with big data, we identified fluorescence features in green pepper images that are imperceptible under natural light. Through mechanistic analysis, we determined that over 93% of these fluorescence features naturally arise during the growth process and remain stable. Leveraging feature point matching algorithms, we constructed a label-free tracking system that enabled the traceability of individual green peppers, achieving an 83% accuracy rate in line with commercial application standards. The integration of the indoor positioning system and fluorescence imaging system aligns with the overarching objective of this research, which is to elevate precision agriculture to a level where plant-to-plant accuracy is attainable. We have authored or co-authored 6 SCI papers for showing the new findings as well as presentations at 5 international academic conferences. These publications and conference contributions play a crucial role in amplifying the impact of our research endeavors.
|
