FlowerPhenoNet: Automated Flower Detection from Multi-View Image Sequences Using Deep Neural Networks for Temporal Plant Phenotyping Analysis

A phenotype is the composite of an observable expression a genome for traits in given environment. The trajectories phenotypes computed from image sequence and timing important events plant’s life cycle can be viewed as temporal indicative growth pattern vigor. In this paper, we introduce novel method called FlowerPhenoNet, which uses deep neural networks detecting flowers multiview sequences high-throughput plant phenotyping analysis. Following flower detection, set flower-based are computed, e.g., day emergence first cycle, total number present at time, highest bloomed plant, trajectory flower, blooming plant. To develop new algorithm facilitate performance evaluation based on experimental analysis, benchmark dataset indispensable. Thus, FlowerPheno, comprises three flowering species, sunflower, coleus, canna, captured by visible light camera platform multiple view angles. analyses FlowerPheno demonstrate efficacy FlowerPhenoNet.