Study on temporal variation in citrus canopy using thermal imaging for citrus fruit detection

This paper presents a study of the thermal temporal variation in citrus canopy as a potential approach for improving fruit detection for harvesting orange. Tree canopy was monitored on 24 h cycles using a thermal infrared camera. Four different trees (4 regions of interest in 4 different tree canopies) were observed on four separate days. Surface temperature of the fruit, ambient temperature and relative humidity (RH) were measured using a portable Dew Point Meter. The acquired thermal images were corrected for fruit emissivity, which was estimated to be 0.90, the ambient temperature, the RH and the reflected temperature. Fruit and canopy temperature profile demonstrated a relatively large temperature gradient, which occurred from afternoon (16:00) until midnight. This was evident from the acquired thermal images of the canopy. In addition, the fruits were successfully segmented in the thermal images using image processing techniques during the time range of the largest temperature difference, which suggest potential application of thermal imaging for improved detection for harvesting.