ocrs

What ocrs is

ocrs is an open-source Rust library and command-line tool for extracting text from images with OCR. It is intended for scanned documents, text-containing photos, and screenshots, with a goal of needing less preprocessing than older engines such as Tesseract. The CLI can save recognized text, return text and layout details as JSON, or create an annotated image showing detected words and lines. An optional clipboard feature enables OCR on images from the system clipboard. Developers can use the Rust library in their applications, or the ocrs-capi component, which provides a C ABI for creating an engine and returning recognized text with line bounding boxes. Its neural models are trained in PyTorch, exported to ONNX, and run using the RTen engine. On first run, the CLI downloads required models to `~/.cache/ocrs`. Installing the CLI requires Rust and Cargo, with the command `cargo install ocrs-cli --locked`. The project describes itself as an early preview and warns that it may make more errors than commercial OCR engines. It currently recognizes the Latin alphabet, while wider language support is planned. The repository lists Apache-2.0 and MIT licenses.