Deep Leaf Segmentation Using Synthetic Data
This work addresses the need for automated leaf segmentation in plant phenotyping, but it is incremental as it applies an existing method (Mask-RCNN) with synthetic data augmentation.
The paper tackled the problem of leaf instance segmentation for high-throughput phenotyping by augmenting real plant datasets with synthetic images, achieving a 90% leaf segmentation score on the A1 test set and 81% mean performance over all five test datasets.
Automated segmentation of individual leaves of a plant in an image is a prerequisite to measure more complex phenotypic traits in high-throughput phenotyping. Applying state-of-the-art machine learning approaches to tackle leaf instance segmentation requires a large amount of manually annotated training data. Currently, the benchmark datasets for leaf segmentation contain only a few hundred labeled training images. In this paper, we propose a framework for leaf instance segmentation by augmenting real plant datasets with generated synthetic images of plants inspired by domain randomisation. We train a state-of-the-art deep learning segmentation architecture (Mask-RCNN) with a combination of real and synthetic images of Arabidopsis plants. Our proposed approach achieves 90% leaf segmentation score on the A1 test set outperforming the-state-of-the-art approaches for the CVPPP Leaf Segmentation Challenge (LSC). Our approach also achieves 81% mean performance over all five test datasets.