How bunyaveedeechuay/automatic-panorama-stitcher works
Panorama Stitcher is a browser-only app: users add and reorder photos, stitch them sequentially, then download a PNG. Its main workflow loads and resizes each image, finds ORB features, matches them, estimates homographies, and warps/composites each pair. The graph reflects documented capabilities and the sampled orchestration; some UI wiring and OpenCV loading details were not present in the excerpts, so uncertain links are omitted.
Main entry points: app.js Sequential Stitching · index.html Browser UI
Components 15
Photo Input
- Browser UI
index.html - Photo Input
app.js - Frame State
app.js - Filmstrip
app.js - Stitch Control
app.js
Stitch Pipeline
- Image Load/Resize
app.js - ORB Features
app.js - Feature Matching
app.js - Homography
app.js - Warp & Composite
app.js - Sequential Stitching
app.js
Result Export
- Output Canvas
index.html - PNG Download
app.js
Other components
- User
- OpenCV.js
Connections 18
- User → Browser UI: uses
- Browser UI → Photo Input: accepts files
- Photo Input → Frame State: adds frames
- Frame State → Filmstrip: renders frames
- User → Filmstrip: reorders frames
- User → Stitch Control: starts stitching
- Stitch Control → Sequential Stitching: invokes
- Sequential Stitching → Image Load/Resize: loads frames
- Image Load/Resize → OpenCV.js: creates Mats
- Sequential Stitching → Warp & Composite: stitches each pair
- Warp & Composite → ORB Features: detects features
- ORB Features → Feature Matching: provides descriptors
- Feature Matching → Homography: provides matches
- Homography → Warp & Composite: provides transform
- Warp & Composite → OpenCV.js: uses CV operations
- Sequential Stitching → Output Canvas: displays result
- Output Canvas → PNG Download: exports PNG
- User → PNG Download: downloads