Plant characteristic estimation using sonar, multispectral reflectance, and electromagnetic response

Abstract

Scope and Method of Study:

The goal of this study was to design, test and validate three methods of remotely estimating plant physical and physiological characteristics. A free-space parallel plate electrostatic sensing system operating at medium radio frequency range was used to estimate water content and plant dry biomass. An ultrasound distance sensing system and a multispectral imaging system was used to directly estimate plant height and top view surface area and indirectly estimate plant biomass. NDVI was calculated from the multispectral imaging system data. Combining NDVI with the plant height and top view surface area estimates, a correlation was observed between plant biomass, chlorophyll content and chlorophyll concentration.

Findings and Conclusions:

Plant water content and dry biomass of greenhouse grown spinach were estimated using a free-space electrostatic sensing system (r2 = 0.95).

Ultrasonic sensor-based height estimates and top view surface area multispectral image data provided plant biomass estimates in corn and spinach (r2 = 0.85 and 0.88). Estimates for snap beans were not as convincing (r2 = 0.52).

Combining biomass estimates from the height and surface area data obtained by the ultrasonic distance sensor and the multispectral imaging system with NDVI670 calculated from reflectance data from the imaging system provided strong correlations with chlorophyll content in spinach (r2 = 0.91). This was an improvement from the chlorophyll content estimates using only NDVI670. Correlations with chlorophyll concentration were weak. The strongest correlation was found using the reflectance ratio, NIR/Green (r2 = 0.30).