Global Rice Multi-Class Segmentation Dataset (RiceSEG): A Comprehensive and Diverse High-Resolution RGB-Annotated Images for the Development and Benchmarking of Rice Segmentation Algorithms

2025-04-02

AI Summary

Field-based rice phenotyping is hindered by the difficulty of fine-grained organ segmentation and the scarcity of high-quality, pixel-level annotations. To address this, we introduce RiceSEG—the first large-scale, cross-national (5 countries), multi-cultivar (6,000+ genotypes), full-growth-cycle semantic segmentation benchmark for rice. It comprises 3,078 high-resolution RGB images with pixel-accurate annotations for six anatomical organs and common confounding objects (e.g., weeds, soil). RiceSEG fills a critical gap in crop-specific segmentation datasets. Using it, we systematically benchmark state-of-the-art models—including DeepLabv3+ and SegFormer—and identify significant performance bottlenecks: notably low mIoU on panicles, senescent leaves, and weeds (average improvement margin ≈22%), especially under complex reproductive-stage canopies. RiceSEG thus establishes a rigorous, community-standard evaluation framework for organ-level rice phenotyping and segmentation algorithm development.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Evaluation and Analysis

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Developing computer vision-based rice phenotyping techniques is crucial for precision field management and accelerating breeding, thereby continuously advancing rice production. Among phenotyping tasks, distinguishing image components is a key prerequisite for characterizing plant growth and development at the organ scale, enabling deeper insights into eco-physiological processes. However, due to the fine structure of rice organs and complex illumination within the canopy, this task remains highly challenging, underscoring the need for a high-quality training dataset. Such datasets are scarce, both due to a lack of large, representative collections of rice field images and the time-intensive nature of annotation. To address this gap, we established the first comprehensive multi-class rice semantic segmentation dataset, RiceSEG. We gathered nearly 50,000 high-resolution, ground-based images from five major rice-growing countries (China, Japan, India, the Philippines, and Tanzania), encompassing over 6,000 genotypes across all growth stages. From these original images, 3,078 representative samples were selected and annotated with six classes (background, green vegetation, senescent vegetation, panicle, weeds, and duckweed) to form the RiceSEG dataset. Notably, the sub-dataset from China spans all major genotypes and rice-growing environments from the northeast to the south. Both state-of-the-art convolutional neural networks and transformer-based semantic segmentation models were used as baselines. While these models perform reasonably well in segmenting background and green vegetation, they face difficulties during the reproductive stage, when canopy structures are more complex and multiple classes are involved. These findings highlight the importance of our dataset for developing specialized segmentation models for rice and other crops.

Problem

• Develop high-quality rice segmentation dataset for phenotyping.
• Address challenges in rice organ segmentation due to complexity.
• Benchmark segmentation models for diverse rice growth stages.

Innovation

• High-resolution RGB images for rice segmentation
• Multi-class semantic segmentation dataset RiceSEG
• CNN and transformer-based baseline models tested