用稀疏标注训练模型，实现香蕉瑕疵的自动分级
Weakly Supervised Panoptic Segmentation for Defect-Based Grading of Fresh Produce
- 用SAM生成稠密分割掩码，减少人工标注量
- 在476张图像上识别出1440个瑕疵，准确率提升显著
- 适合数据少、人力成本高的农产品质检场景
农产品供应链中的视觉缺陷分级至关重要，但传统方式依赖人工且易出错。自动化计算机视觉方法通常需要大量标注数据，而这类数据在分散式供应链中难以获取。本文评估了通用分割模型SAM，通过稀疏标注生成密集的全景分割掩码，并用于训练监督式全景分割模型。以香蕉表面瑕疵（淤青和疤痕）为例，基于476张实地采集图像及1440个标注瑕疵进行验证。尽管SAM生成的掩码整体与人工标注吻合，显著降低标注成本，但仍存在特定尺寸和形状瑕疵的失败案例。不过，该方法仍能从全景掩码中获得瑕疵数量和相对大小的可靠估计，揭示了基础模型在低数据农业场景下缺陷量化中的潜力与局限。
原文摘要 · Abstract (English)
Visual inspection for defect grading in agricultural supply chains is crucial but traditionally labor-intensive and error-prone. Automated computer vision methods typically require extensively annotated datasets, which are often unavailable in decentralized supply chains. We address this challenge by evaluating the Segment Anything Model (SAM) to generate dense panoptic segmentation masks from sparse annotations. These dense predictions are then used to train a supervised panoptic segmentation model. Focusing on banana surface defects (bruises and scars), we validate our approach using 476 field images annotated with 1440 defects. While SAM-generated masks generally align with human annotations, substantially reducing annotation effort, we explicitly identify failure cases associated with specific defect sizes and shapes. Despite these limitations, our approach offers practical estimates of defect number and relative size from panoptic masks, underscoring the potential and current boundaries of foundation models for defect quantification in low-data agricultural scenarios. GitHub: https://github.com/manuelknott/banana-defect-segmentation
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。