Pre-symptomatic disease detection in the vine, chrysanthemum, and rose leaves with a low-cost infrared sensor

Abstract

Thermography is a technique based on infrared imaging, which is used nowadays to detect plants under stress caused by biotic and abiotic factors. In many cases, temperature changes have already been correlated with pathogen attacks. In this sense, thermography offers the ability of early disease detection in plant pathology. In this work, a low-cost AMG8833 Grind-Eye infrared camera combined with a 1080P RGB web camera was used to develop an integrated infrared and RGB imaging system, to record temperature changes on vine, chrysanthemum, and rose plant leaf surfaces. Vine and chrysanthemum leaves were infected with Phomopsis viticola and Septoria ssp. respectively, respectively, whereas rose plants leaves were infected with Colletotrichum spp. as well as with Podosphaera pannosa. Measurements were performed using the integrated imaging system on infected and uninfected leaves, as well as on PDA plates with active and non-active mycelium. According to the results, vine leaf tissue infected with P. viticola and rose plants leaf tissue infected with P. pannosa had a pre-symptomatic (four days after infection) decrease in temperature up to 1.6 and 1.1◦C, respectively, compared with uninfected tissue. In contrast chrysanthemum leaf tissue infected with Septoria ssp. and rose plant leaf tissue infected with Colletotrichum spp. had a pre-symptomatic (four days after infection) increased temperature up to 1.1◦C and 1.0◦C, respectively, compared with uninfected tissue. In vitro measurements showed that the active fungi mycelium had approximately 1.1 to 2.1◦C lower temperature than the non-active mycelium. The results above show that the integrated infrared and RGB imaging system developed in this work can be used to detect early disease infection before visible symptoms appeared, facilitating the decision-making process.