How harshith-reddy11/remote-sensing-image-analysis-using-deep-learning works
This is a Tkinter desktop application for remote-sensing image analysis. The documented workflow is dataset upload and preprocessing, feature/model analysis, then prediction and evaluation. The sampled Main.py implements image/annotation loading, train/test splitting, a custom CNN with bounding-box and class outputs, prediction, and training-graph display; the README additionally describes ResNeXt-50 and LSTM, which are not verified in the excerpt. The repository is very small, so these responsibilities share one source path. Connections are limited to relationships visible in the excerpt or documented in the README.
Main entry points: Main.py Tkinter interface
Components 11
Desktop workflow
- Tkinter interface
Main.py
Image analysis
- Dataset loading
Main.py - Dataset preprocessing
Main.py - CNN detection
Main.py - Image prediction
Main.py - Metrics and plots
Main.py
Local data and models
- Processed data cache
Main.py - Model weights
Main.py
Other components
- Analyst
- RSSOD images
- Displayed results
Connections 14
- Analyst → Tkinter interface: uses
Main.py - Tkinter interface → Dataset loading: starts upload
Main.py - Dataset loading → RSSOD images: reads images
Main.py - Dataset loading → Processed data cache: loads or saves
Main.py - Tkinter interface → Dataset preprocessing: starts split
Main.py - Dataset preprocessing → CNN detection: provides splits
Main.py - Tkinter interface → CNN detection: starts training
Main.py - CNN detection → Model weights: saves or loads
Main.py - CNN detection → Displayed results: shows predictions
Main.py - Tkinter interface → Image prediction: selects image
Main.py - Image prediction → Model weights: loads model
Main.py - Image prediction → Displayed results: draws detections
Main.py - Tkinter interface → Metrics and plots: opens graph
Main.py - Metrics and plots → Displayed results: displays metrics
Main.py