RapidX annotator: A specialized software tool for industrial radiographic image annotation and enhancement

• A user-friendly tool for industrial radiographic image annotation and enhancement. • Integrates YOLO-based automatic pre-annotation with manual refinement functions. • Supports local contrast enhancement, denoising, and pseudo-color conversion. • Exports annotation data in standard XML and JSON formats for AI model training. • Enables efficient labeling of complex grayscale radiographic images. Radiographic testing (RT) plays a vital role in non-destructive evaluation (NDT), yet the interpretation and annotation of radiographic images remain time-consuming and error-prone. To address this, we introduce RapidX Annotator, an open-source software tool tailored for industrial radiographic image annotation and enhancement. The software combines advanced image processing capabilities—including localized contrast adjustment, denoising, and pseudo-color conversion—with geometric transformations and versatile annotation tools. A key feature is the integration of YOLO-based deep learning for automated pre-annotation, enabling users to refine predictions efficiently. Validation on the SWRD dataset demonstrated improved defect visibility and a 35.9 % reduction in annotation time. RapidX Annotator outputs in standard XML/JSON formats and supports cross-platform deployment. This tool bridges the gap between manual annotation and AI-assisted defect detection, promoting faster, more accurate data preparation for training robust inspection models in NDT workflows.