Smart Grocery Monitoring System

• 

Inspiration:

I was inspired by the growing need for automated inventory management in retail stores. Often, shelves remain empty or products get misplaced, leading to lost sales and customer dissatisfaction. I wanted to build a real-time AI system that could monitor shelves and alert managers, reducing manual effort and improving efficiency.

What it does

Fully functional AI-powered shelf monitoring system.

Supports live CCTV feeds and video/image uploads.

Generates real-time dashboards, alerts, and reports.

Reduces manual effort, improves stock management, and provides insights for store managers.

How we built it:

Data & Model: Used YOLOv8 pre-trained weights to detect retail products.

Detection Pipeline:

results = model(frame, stream=True) for r in results: for box in r.boxes: cls = int(box.cls[0]) label = model.names[cls]

Dashboard: Streamlit with 3 columns:

Live video feed

Product count table

Interactive charts (bar, pie, trend)

Alerts: Integrated SMTP emails to notify low-stock products.

Export Reports: Automatically generated CSV/Excel reports of detected stock.

Challenges we ran into

Real-Time Processing: Handling live video without lag required efficient frame-by-frame detection.

File Upload Limits: Streamlit default 200MB limit had to be managed via config file.

Multiple Formats: Supporting both video and image uploads required conditional handling.

Dynamic Dashboard: Keeping charts interactive and real-time was tricky with live streaming.

Accomplishments that we're proud of

We successfully built a real-time AI-powered shelf monitoring system that detects and counts products, generates interactive dashboards, sends low-stock alerts, and supports both live CCTV feeds and video uploads — all fully automated and optimized for real-world retail environments.

What we learned:

Computer Vision & Object Detection: Learned to use YOLOv8 for detecting multiple products in real-time.

Real-Time Data Processing: Managed live video feeds efficiently with OpenCV.

Interactive Dashboards: Built responsive dashboards using Streamlit, including bar, pie, and trend charts.

Automation & Alerts: Implemented smart email alerts for low or out-of-stock items.

Data Handling & Analysis: Used Pandas for counting and aggregating product data.

What's next for Smart Grocery Monitoring System:

Built With

• amazon-web-services
• compatible-with-cctv/webcam-streams-optional-cloud-/-deployment:-can-be-deployed-on-streamlit-cloud
• excel-export-via-pandas-platforms:-local-machine
• for
• gcp
• numpy-data-handling-&-analysis:-pandas-web-/-dashboard-framework:-streamlit-(interactive-dashboards)-visualization:-plotly-(bar
• or
• pie
• programming-languages:-python-ai-/-machine-learning-frameworks:-yolov8-(ultralytics)-for-object-detection-computer-vision:-opencv
• python-smtplib-file-handling-&-reporting:-csv
• real-time
• trend-charts)-email-alerts:-smtp