Fig. 1From: LeTra: a leaf tracking workflow based on convolutional neural networks and intersection over unionGraphical abstract of the proposed approach, consisting of two main stages. In the first stage (A), a detection and segmentation model is trained using images of A. thaliana rosettes. In the second stage (B), the trained model is employed to detect and segment individual images at each time step. The resulting masks are ranked according to their IoU as denoted by green and red arrows representing high and low IoU values, respectively. This ranking enables the association and reconstruction of object trackingBack to article page