How this analysis was created: This summary and feature list are AI-generated from collected project material and can contain mistakes. Stars, license and language are imported from GitHub. Inclusion does not mean that we have tested or audited this project. Check the source documentation for any feature you depend on. Learn more on our About page.
[x] 支持GPU，CPU（opencv dnn加速）； - [ ] 整合darknet-ocr完成对表格的重建，输出json\excel
The main features of chineseocr/table-ocr are: Table Processing.
Projects with overlapping indexed features include: atlanhq/camelot — Camelot is a Python-based library designed to parse, extract, and clean tabular data from PDF files. It converts table… carefree0910/carefree-learn — Deep Learning ❤️ PyTorch. chezou/tabula-py — Simple wrapper of tabula-java: extract table from PDF into pandas DataFrame. diyago/gan-for-tabular-data — Generative Networks are well-known for their success in realistic image generation. However, they can also be applied… google-research/tapas — End-to-end neural table-text understanding models. holms-ur/fine-tuning — Close-Domain fine-tuning for table detection.
Simple wrapper of tabula-java: extract table from PDF into pandas DataFrame
Generative Networks are well-known for their success in realistic image generation. However, they can also be applied to generate tabular data. We introduce major improvements for generating high-fidelity tabular data giving oppotunity to try GANS, TimeGANs, Diffusions and LLM for tabular data…
Camelot is a Python-based library designed to parse, extract, and clean tabular data from PDF files. It converts table elements from text-based PDF documents into programmable data structures and dataframes. The tool identifies tabular regions using coordinate-based grouping, lattice-based line detection, and stream-based text extraction. It can also rasterize PDF pages into images to utilize computer vision for detecting structural lines and boundaries. Extracted data is validated through accuracy and whitespace metrics to filter out low-quality extractions. The processed information can be
