Get Free Synopsis PDF
We will prepare a detailed synopsis for your college submission. Includes problem statement, objectives, DFD, ER diagram, and more.
A simple web app that converts photos of handwritten digits and notes into editable text using a convolutional neural network trained on public handwriting datasets.
Problem Statement
Typing up handwritten notes, forms, and answer sheets is slow and error prone. Students need an approachable project that covers the full ML workflow end to end.
Objectives
Train a CNN on MNIST and EMNIST datasets
Preprocess uploaded images with OpenCV
Segment lines and characters before recognition
Serve predictions through a web interface
Modules
Model Training
CNN training notebook with accuracy and confusion matrix reports.
Image Preprocessing
Thresholding, noise removal, and character segmentation.
Recognition API
Flask endpoint that returns recognized text for an uploaded image.
Web UI
Upload or draw on a canvas and copy or download the recognized text.
Expected Output
A web page where the user uploads a handwritten image or draws on a canvas and sees the recognized text instantly.
Future Scope
Cursive handwriting support with sequence models, Hindi and regional scripts, and a mobile camera scanner.