Shape and Style GAN-based Multispectral Data Augmentation for Crop/Weed Segmentation in Precision Farming

TL;DR

This paper introduces a novel multispectral data augmentation method using Shape and Style GANs, enhancing crop and weed segmentation in precision farming by generating high-quality artificial images that improve training datasets.

Contribution

It presents a GAN-based augmentation technique that replaces object patches in original images with synthetic ones, improving data diversity for crop/weed segmentation.

Findings

Enhanced segmentation accuracy on public datasets.

Effective augmentation with high-quality synthetic images.

Open source code and data availability.

Abstract

The use of deep learning methods for precision farming is gaining increasing interest. However, collecting training data in this application field is particularly challenging and costly due to the need of acquiring information during the different growing stages of the cultivation of interest. In this paper, we present a method for data augmentation that uses two GANs to create artificial images to augment the training data. To obtain a higher image quality, instead of re-creating the entire scene, we take original images and replace only the patches containing objects of interest with artificial ones containing new objects with different shapes and styles. In doing this, we take into account both the foreground (i.e., crop samples) and the background (i.e., the soil) of the patches. Quantitative experiments, conducted on publicly available datasets, demonstrate the effectiveness of the proposed approach. The source code and data discussed in this work are available as open source.