New OCR pipeline enhances retail bill digitization with adaptive enhancement

Researchers have developed and benchmarked an adaptive Optical Character Recognition (OCR) pipeline specifically designed for digitizing diverse retail bills. This system incorporates a CNN-based enhancement module, an image quality analyzer, and an NLP-based correction layer to handle variations in scan quality and layout. The proposed pipeline demonstrated significant improvements over the Tesseract baseline, achieving a Character Error Rate of 18.4% and a Word Error Rate of 27.6% on a dataset of 360 retail bill images. AI

影响 Establishes a new benchmark for OCR in retail bill digitization, potentially improving efficiency for businesses dealing with varied document formats.

排序理由 This is a research paper detailing a new OCR pipeline and its benchmark results.

在 Hugging Face Daily Papers 阅读 →

• Convolutional Neural Network
• Laplacian variance
• Tesseract

• 论文
• 其他

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →