Enhanced Droplet Analysis Using Generative Adversarial Networks

TL;DR

This work tackles data scarcity in precision agriculture droplet analysis by introducing DropletGAN, a progressive GAN that synthesizes high-resolution droplet images from a small real dataset. The model grows from 4×4 to 1024×1024, achieving realistic droplets with an $FID$ of 11.29, and pairs with a YOLOv8-based detector trained on synthetic data to boost detection performance by about 16 percentage points in mean average precision. The synthetic data enable notable improvements in droplet detection metrics, demonstrating the viability of synthetic augmentation to reduce costly data collection for nozzle design and spray-automation tasks. Overall, DropletGAN offers a resource-efficient path to enhance precision spraying and supports broader agricultural AI applications by mitigating data-collection bottlenecks.

Abstract

Precision devices play an important role in enhancing production quality and productivity in agricultural systems. Therefore, the optimization of these devices is essential in precision agriculture. Recently, with the advancements of deep learning, there have been several studies aiming to harness its capabilities for improving spray system performance. However, the effectiveness of these methods heavily depends on the size of the training dataset, which is expensive and time-consuming to collect. To address the challenge of insufficient training samples, we developed an image generator named DropletGAN to generate images of droplets. The DropletGAN model is trained by using a small dataset captured by a high-speed camera and capable of generating images with progressively increasing resolution. The results demonstrate that the model can generate high-quality images with the size of 1024x1024. The generated images from the DropletGAN are evaluated using the Fréchet inception distance (FID) with an FID score of 11.29. Furthermore, this research leverages recent advancements in computer vision and deep learning to develop a light droplet detector using the synthetic dataset. As a result, the detection model achieves a 16.06% increase in mean average precision (mAP) when utilizing the synthetic dataset. To the best of our knowledge, this work stands as the first to employ a generative model for augmenting droplet detection. Its significance lies not only in optimizing nozzle design for constructing efficient spray systems but also in addressing the common challenge of insufficient data in various precision agriculture tasks. This work offers a critical contribution to conserving resources while striving for optimal and sustainable agricultural practices.

Figures (10)

• Figure 1: Application of spray systems in agriculture (image is taken from www.grainews.ca)
• Figure 2: Experimental setup of a spray system for droplet generation.
• Figure 3: Droplet images captured from a high-speed camera in the experiments.
• Figure 4: The architecture of the GAN model. The generator and discriminator are trained together to finalize the best generator.
• Figure 5: The architecture of the generative model in the training process.