Deep Leaf Segmentation Using Synthetic Data

TL;DR

This paper introduces a synthetic data augmentation framework for leaf segmentation that enhances deep learning performance, achieving state-of-the-art results on benchmark datasets by combining real and synthetic images.

Contribution

The paper presents a novel approach using synthetic data generated through domain randomisation to improve deep learning-based leaf segmentation.

Findings

Achieved 90% segmentation score on the A1 test set.

Outperformed existing methods on the CVPPP Leaf Segmentation Challenge.

Achieved 81% mean performance across five datasets.

Abstract

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.