How vanisri6396-hue/industrialvision_ai works
IndustrialVision AI is a YOLO-based industrial object-detection project. Its implemented user-facing path is image upload in the Streamlit app, inference through the shared detection module, and display/download of annotated results. Separate scripts provide video detection, webcam inference, object tracking, model evaluation, and PPE fine-tuning with dataset checks. The app’s video, webcam, tracking, and analytics pages are placeholders; no analytics store or safety-alert implementation is evidenced. Standalone workflow details are based on sampled scripts.
Main entry points: app.py Streamlit UI · detection.py Image Inference · image_detection.py Image Script · video_detection.py Video Detection
Components 12
Streamlit Application
- Streamlit UI
app.py
Detection Workflows
- Image Inference
modules/detection.py - Image Script
image_detection.py - Video Detection
video_detection.py - Video Tracking
tracking_video.py - Webcam Detection
webcam_detection.py
Model Operations
- Model Evaluation
evaluate_model.py - PPE Fine-tuning
train_ppe.py
Dataset Quality
- Dataset Validation
validate_dataset.py
Other components
- Operator
- Ultralytics YOLO
- Detection Results
Connections 10
- Operator → Streamlit UI: uses
app.py - Streamlit UI → Image Inference: calls
app.py - Image Inference → Ultralytics YOLO: loads and predicts
modules/detection.py - Streamlit UI → Detection Results: displays and exports
app.py - Image Script → Ultralytics YOLO: predicts
image_detection.py - Video Detection → Ultralytics YOLO: predicts frames
video_detection.py - Video Tracking → Ultralytics YOLO: tracks frames
tracking_video.py - Webcam Detection → Ultralytics YOLO: predicts frames
webcam_detection.py - Model Evaluation → Ultralytics YOLO: validates model
evaluate_model.py - PPE Fine-tuning → Ultralytics YOLO: trains and validates
train_ppe.py