评测4种开源OCR在南非食品包装上的表现，为标签识别提供基准。
Evaluating OCR performance on food packaging labels in South Africa
- 对比Tesseract、EasyOCR等4个OCR模型在真实包装图上的表现。
- Tesseract在字符错误率上最低（0.912），PaddleOCR覆盖最全但速度慢。
- 结果揭示布局感知与文本定位对包装识别的关键作用，适合营养分析研究者。
本研究评估了四种开源光学字符识别（OCR）系统——Tesseract、EasyOCR、PaddleOCR和TrOCR——在真实世界食品包装图像上的表现，重点考察其提取成分表和营养信息面板的能力。准确的包装OCR对合规性检查和营养监测至关重要，但因多语言文本、密集排版、字体多样、反光及曲面等因素而极具挑战。研究使用包含231种产品（共1,628张图像）的数据集测试各模型的速度与覆盖率，并构建了由113张图像（60种产品）组成的真值子集用于精度评估。评估指标包括字符错误率（CER）、词错误率（WER）、BLEU、ROUGE-L、F1、覆盖率和执行时间。在真值子集上，Tesseract达到最低的CER（0.912）和最高的BLEU（0.245）。EasyOCR在准确率与多语言支持之间取得良好平衡。PaddleOCR实现近乎完整的覆盖率，但因仅在CPU运行（GPU不兼容）导致速度较慢；TrOCR虽经GPU加速，结果却最差。研究结果建立了面向包装场景的基准，提供了性能基线，并指明布局感知方法与文本定位技术的发展方向。
原文摘要 · Abstract (English)
This study evaluates four open-source Optical Character Recognition (OCR) systems which are Tesseract, EasyOCR, PaddleOCR, and TrOCR on real world food packaging images. The aim is to assess their ability to extract ingredient lists and nutrition facts panels. Accurate OCR for packaging is important for compliance and nutrition monitoring but is challenging due to multilingual text, dense layouts, varied fonts, glare, and curved surfaces. A dataset of 231 products (1,628 images) was processed by all four models to assess speed and coverage, and a ground truth subset of 113 images (60 products) was created for accuracy evaluation. Metrics include Character Error Rate (CER), Word Error Rate (WER), BLEU, ROUGE-L, F1, coverage, and execution time. On the ground truth subset, Tesseract achieved the lowest CER (0.912) and the highest BLEU (0.245). EasyOCR provided a good balance between accuracy and multilingual support. PaddleOCR achieved near complete coverage but was slower because it ran on CPU only due to GPU incompatibility, and TrOCR produced the weakest results despite GPU acceleration. These results provide a packaging-specific benchmark, establish a baseline, and highlight directions for layout-aware methods and text localization.
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。