Infrared imaging indices for genotype screening in plant drought responses

Abstract

The crucial obstacle in the research of plant science is to distinguish between superior genotypes and its selection. Machine vision can recognize the better genotype precisely, effortlessly and in asymptomatic manner. Genotype differentiation based on machine vision by using thermal imaging parameters is evaluated and elaborated in this article. The main objective of this study is to assess better performance and select superior genotypes based on the common thermal indicators. Mungbean and chickpea crops were studied and measured in greenhouse conditions with well-watered and water stress treatments. An algorithm is developed for extracting parameters from thermal images and implemented in a Python tool using OpenCV (cv2), Pandas packages. The genotypes of the crops were contrasted for drought tolerance to be able to differentiate drought responses with thermal imaging. The result of the experiments express that crops are differentiated between treatments and discriminated among genotypes within a treatment. These results were validated with the soil moisture data, which was collected on simultaneous day of the image captured.