How enricocasaquia/reconhecimento_incendio works
This is a command-line fire-image classifier with two distinct result paths: CNN training/loading and image inference, plus a separate color-based outline view and training-metric plots. The README describes training, loading, batch testing, outlines, and model summaries. The map reflects the sampled control flow; the saved model artifact is included as a stateful input. Dataset and sample contents were not inspected, so no runtime links to them are asserted.
Main entry points: main.py CLI orchestration · FireDetection.py Model training · FireImageOutliner.py Outline workflow · modelo_deteccao_fogo.keras Trained model
Components 18
Interactive workflow
- CLI orchestration
main.py - Single image test
main.py - Directory test
main.py
CNN lifecycle
- Model training
tools/FireDetection.py - CNN architecture
tools/FireDetection.py - Dataset iterators
tools/FireDetection.py - Trained model
models/modelo_deteccao_fogo.keras - Image inference
tools/FireDetection.py - Model summary
tools/FireDetection.py - Validation evaluation
tools/FireDetection.py
Image analysis
- Outline workflow
tools/FireImageOutliner.py - HSV filtering
tools/FireImageOutliner.py - Fire mask
tools/FireImageOutliner.py - Contour detection
tools/FireImageOutliner.py - Outline display
tools/FireImageOutliner.py
Visualization
- Training metrics
tools/FireDetection.py - Metric graphs
tools/Graph.py
Other components
- Operator
Connections 21
- Operator → CLI orchestration: provides choices
main.py - CLI orchestration → Model training: starts training
main.py - CLI orchestration → Image inference: loads model
main.py - CLI orchestration → Single image test: routes image test
main.py - CLI orchestration → Directory test: routes directory test
main.py - CLI orchestration → Outline workflow: starts outline
main.py - CLI orchestration → Model summary: requests summary
main.py - CLI orchestration → Outline display: shows outline
main.py - Single image test → Image inference: tests image
main.py - Directory test → Image inference: tests images
main.py - Model training → CNN architecture: constructs
tools/FireDetection.py - Model training → Dataset iterators: creates iterators
tools/FireDetection.py - Model training → Trained model: saves model
tools/FireDetection.py - Model training → Validation evaluation: evaluates
tools/FireDetection.py - Image inference → Trained model: loads or predicts
- Model training → Training metrics: records history
tools/FireDetection.py - Training metrics → Metric graphs: creates plots
tools/FireDetection.py - Outline workflow → HSV filtering: filters image
tools/FireImageOutliner.py - Outline workflow → Fire mask: creates mask
tools/FireImageOutliner.py - Outline workflow → Contour detection: draws contours
tools/FireImageOutliner.py - Outline workflow → Outline display: shows result
tools/FireImageOutliner.py