This is Clojure project for recognizing images of handwritten characters representing the digits 0-9
The code is broken down into the following sections:
*image_access.clj:
- this is the vision module for analyzing an image of a block of text, identifying the pixel coordinates where characters lay in the image, accessing the locations of those characters, re-sizing them to square blocks that in the proper dimensions that the neural network is configured to receive, and processing those image blocks to prepare a vectorized representation each character of it's lightness or darkness score on a scale from 0-1. *neural.clj:
- this is the main file for defining the neural networks processing and training algorithms *initial.clj:
- this is where the neural net reads networks settings from for initializing before training or for processing an input *resources/training-log:
- update the destination in the neural.clj file when beginning training *resources:
- training data, sample images to test with
To train the network you need to update neural.clj and then you can lein run the project.
A live version where you can upload an image to be analyzed will is coming soon. Depends on successful interfacing of neural.clj and image_access.clj
Copyright © 2014
"THE BEER-WARE LICENSE": zachmarkin@gmail.com wrote this project. As long as you retain this notice you can do whatever you want with this stuff. If we meet some day, and you think this stuff is worth it, you can buy me a beer in return, Zachary Markin