This bachelor’s thesis focuses on the design and implementation of a web application for stitching partial photographs of a large document into a single high-resolution image. The
goal is to automatically align and merge overlapping images while preserving detail, eliminating visible seams, and unifying the lighting across individual photographs — all through
a clear, responsive, and user-friendly web interface without requiring local software installation. The application is based on a client-server architecture using FastAPI with asynchronous SQLAlchemy on the backend and Vue.js on the frontend. For the stitching
computation, an external CLI tool called Exposea is integrated, performing image merging.
Results are stored in AWS S3 and served as a tile pyramid in an interactive Leaflet-based
viewer in the browser. The system supports multi-user collaboration with role-based access
control, photo upload with preview generation, and full job lifecycle management including re-runs and failure log access. The result is a fully functional web platform enabling
users to manage projects, upload images, configure and submit stitching jobs, monitor their progress in real time, and inspect results in an interactive tile viewer with full-resolution
download support. The contribution of this work lies in delivering a practical architecture
for web applications that orchestrate computationally intensive external tools through a
message queue, making complex image processing pipelines accessible via a standard web browser.