paddleocr-text-recognition

You can extract text from images and PDFs by invoking the provided Python API or CLI wrapper, which communicates with the PaddleOCR API. It processes invoices, forms, and general documents, returning a structured envelope with the full text, page-level positional data, and raw API response.

Yes, performing OCR on invoices and forms requires setting the PADDLEOCR_OCR_API_URL and PADDLEOCR_ACCESS_TOKEN environment variables, with an optional timeout setting. These credentials authenticate your requests to the PaddleOCR API to successfully extract text and layout information.

Yes, you can process multiple PDFs for text extraction using the provided CLI wrapper designed for batch or on-demand OCR tasks. It returns page-wise results and full text for each document, making it simple to automate text extraction workflows across various files.

The structured envelope returned when extracting text from images includes ok, text, result, and error fields. This format ensures you receive the full extracted text, detailed page-level positional data from the PaddleOCR API, and any error messages encountered during processing.

Yes, PaddleOCR text extraction integrates with Python and uses the httpx dependency to handle API communication. This setup allows you to send images and PDFs to the configured PaddleOCR API URL and receive structured text and layout data in return.