Automated leaf movement tracking in time-lapse imaging for plant phenotyping

Abstract The analysis of the rhythm of leaf movement is a simple yet effective method to quantify the impacts of external (e.g. abiotic stress) and/or internal (e.g. gene mutations) perturbations on plant growth. We developed an automated monitoring system to quantify leaf movement using time-lapse imaging and a subsequent leaf-tracking algorithm. The leaf-tracking algorithm was based on dense optical flow algorithm to directly record temporal motion events. The algorithm measures motion directly, rather than detecting leaf or cotyledon tip in every image, so multiple leaves, including occluded leaves, can be measured simultaneously. To test the monitoring system, wild-type and drought-tolerant mutant genotypes of Arabidopsis (Arabidopsis thaliana) were subjected to a combinatorial two water and two nitrogen levels. High-frequency time-lapse images were acquired from top view for little over 6 consecutive days at a frequency of 4 min. Results showed that nitrogen and water treatments elicited differences in mean plant displacement in both genotypes. It also showed significant differences among the two different genotypes in the mean displacement when plants were under water or nitrogen stress. These results confirmed the new monitoring system’s ability to discern environmental and genotypic differences in plant response.