How boggiomichael/syrup works
Syrup is a Rust library for turning image frames into visual observations, not a domain-specific application. A consumer supplies frames (or captures a Windows window), then calls the relevant geometry, motion, OCR, or quality APIs; results and detector metadata are interpreted by that consumer. The README documents these capabilities, while excerpts show motion tracking and shared result types. Some integration details are unsampled, so the graph omits uncertain links rather than assuming a single pipeline.
Main entry points: motion.rs Motion detection · geometry.rs Geometry analysis · mod.rs OCR orchestration · capture.rs Window capture
Components 14
Frame input
- RGBA frames
- Window capture
src/capture.rs
Visual primitives
- Geometry analysis
src/geometry.rs - Color predicates
src/color.rs - Motion detection
src/motion.rs - Centroid tracking
src/tracking.rs
Text recognition
- Text quality
src/quality.rs - OCR orchestration
src/ocr/mod.rs - Windows OCR
src/ocr/windows.rs - Tesseract
Shared support
- Detection results
src/detection.rs - Debug drawing
src/draw.rs - Timing metrics
src/timing.rs
Other components
- Library consumer
Connections 18
- Library consumer → Window capture: requests capture
README.md - Window capture → RGBA frames: returns frames
README.md - Library consumer → RGBA frames: supplies frames
README.md - Library consumer → Geometry analysis: locates regions
README.md - Library consumer → Color predicates: tests pixels
README.md - Library consumer → Motion detection: detects movement
README.md - Library consumer → Text quality: assesses sharpness
README.md - Library consumer → OCR orchestration: recognizes text
README.md - Library consumer → Debug drawing: annotates frames
README.md - Library consumer → Timing metrics: measures timing
README.md - Geometry analysis → Color predicates: uses predicates
src/geometry.rs - Motion detection → Geometry analysis: extracts regions
src/motion.rs - Motion detection → Centroid tracking: updates tracks
src/motion.rs - Motion detection → Detection results: builds results
src/motion.rs - Centroid tracking → Detection results: uses confidence
src/tracking.rs - Text quality → Geometry analysis: uses rectangles
src/quality.rs - OCR orchestration → Windows OCR: offers backend
src/ocr/mod.rs - OCR orchestration → Tesseract: runs OCR
README.md