A Large-Scale In-the-wild Dataset for Plant Disease Segmentation.

Area of Science:

• Agricultural Science
• Computer Vision
• Machine Learning

Background:

• Plant diseases significantly threaten global agriculture and crop yields.
• Accurate diagnosis and treatment rely on robust image segmentation models.
• Existing datasets lack detailed segmentation labels and in-the-wild images, limiting model development.

Purpose of the Study:

• To introduce PlantSeg, a large-scale segmentation dataset for plant diseases.
• To address the limitations of existing datasets by providing in-the-wild images and detailed segmentation masks.
• To establish a unified benchmarking platform for plant disease segmentation algorithms.

Main Methods:

• Collected and annotated a large-scale dataset of plant disease images.
• Focused on in-the-wild images to capture real-world complexity.
• Included detailed, high-quality disease area segmentation masks for each image.

Main Results:

• Established PlantSeg, comprising 7,774 diseased plant images with segmentation masks.
• The dataset features the largest collection of in-the-wild plant disease images with annotations.
• PlantSeg provides a distinct advantage over existing datasets due to annotation type, image source, and scale.

Conclusions:

• PlantSeg serves as a crucial resource for advancing plant disease segmentation research.
• The dataset facilitates the development of more robust and accurate AI models for agricultural applications.
• This work supports improved crop protection strategies through enhanced disease detection.