We investigated a method to accurately determine plant height spatially from a top view of paprika plants (cultivar Fairway) using a low-cost 3D Kinect from Microsoft. Comparison of plant height between measured values and calculated values (using four types of algorithms) showed that averaging the distance of the top 10 % of pixels closer to the Kinect and then multiplying by a factor of 1.1 yielded the most accurate data. In addition, we developed an algorithm that automatically tracked the shoot apex of plants to measure individual plant and showed that plant height can be measured with high accuracy. The coefficient of determination (R²) was 0.99 or higher, when the calculated and measured values were compared. In addition, it was possible to deduce that humidity deficit reduced plant growth when meteorological data were correlated to changes in plant height recorded by Kinect. The results indicate that changes in plant height recorded by Kinect could be a useful tool for diagnosing plant growth environments. When measurements were recorded at 10-min intervals at night, we observed rapid increase in plant height due to sleeping movement, specifically, an hour after sunset. Therefore, it is advisable to ensure that plants are free of sleeping movement effects for the accurate measurement of plant height.

View full abstract