Turning the old-school hustle of attendance into a seamless digital symphony.
This project automates attendance tracking by recognizing handwritten roll numbers from attendance sheets using OCR technology. It merges the reliability of classic methods with cutting-edge AI to save time, cut errors, and modernize attendance management.
Manual attendance taking? That’s vintage but inefficient. This system steps in to digitize the process by:
- Using EasyOCR to read handwritten digits (roll numbers) from scanned or snapped attendance sheets.
- Offering a smooth user interface for quick image uploads.
- Applying smart preprocessing filters to sharpen image quality before OCR.
- Handling errors gracefully by flagging unclear inputs and guiding users to improve accuracy.
The result? A faster, more reliable attendance system that respects the tradition of roll calls but supercharges it with AI-powered automation.
-
OCR Integration with EasyOCR
Powerful digit extraction from varied handwriting styles, making roll number recognition precise. Supports multiple handwriting quirks—because no two pens write alike. -
User-Friendly Image Upload
A clean, simple frontend lets users upload attendance sheet photos with zero hassle. Backend instantly processes images and returns detected roll numbers for attendance marking. -
Preprocessing Pipeline for Enhanced Accuracy
- Converts images to grayscale for consistency.
- Applies noise reduction filters to clear up smudges and blurs.
- Uses thresholding techniques to highlight digits, boosting recognition success—even with faint or messy handwriting.
-
Robust Error Handling
Automatically flags unclear or ambiguous digits to prevent false attendance records. Provides user feedback on poor image quality or unreadable digits, encouraging re-uploads for clearer images and saving manual correction time.
-
Frontend:
React.js with React Router DOM for multi-page navigation and modular UI experience.
(Includes Welcome, Help, Know More, Attendance Dashboard, Mark Attendance, Monthly Summarizer, AI Anomaly Detection, Developers pages) -
Backend:
Python Flask API handling file uploads, image preprocessing, OCR extraction, and attendance analysis.
Flask-CORS enables smooth cross-origin communication with the frontend. -
OCR Engine:
EasyOCR powered by PyTorch — robust handwritten digit recognition across diverse handwriting styles. -
Image Processing:
OpenCV for advanced image filtering (grayscale, Gaussian blur, adaptive thresholding) to prep images for OCR.
PIL (Python Imaging Library) optionally used in backend. -
Data Handling & Communication:
JSON API responses, enabling seamless frontend-backend data exchange for all core endpoints (/upload,/monthly-summary,/ai-anomaly).
- Upload Image: User uploads a photo or scan of an attendance sheet.
- Preprocess Image: Backend runs filters to optimize image quality.
- OCR Processing: EasyOCR reads and extracts roll numbers.
- Error Checking: System verifies clarity; unclear digits get flagged.
- Result Delivery: Recognized roll numbers are sent back to frontend for display.
- Attendance Marking: User confirms or corrects results before final submission.
- Open the app in your browser.
- Navigate to the attendance upload page.
- Upload a clear photo of the attendance sheet.
- Wait for the backend to process and display roll numbers.
- Review detected digits; fix flagged errors if any.
- Submit attendance digitally with confidence.
Traditional attendance is slow, error-prone, and a drain on time. By marrying time-tested practices with AI-driven OCR, this project honors the past’s structure while embracing the future’s speed and accuracy. It’s about working smarter, not harder, without losing the essence of accountability.
Feel free to reach out for questions or contributions!