CNN to solve image and audio captch using Keras. Build extending the captcha breaker demo by Prof. Andrew Anderson
Required dependencies: python-captcha, opencv, python-tensorflow (CPU or GPU)
./generate.py --width 128 --height 64 --length 4 --symbols symbols.txt --count 3200 --output-dir test
This generates 3200 128x64 pixel captchas with 4 symbols per captcha, using the
set of symbols in the symbols.txt file. The captchas are stored in the folder
test, which is created if it doesn't exist. The names of the captcha images
are scrambled.
To train and validate a neural network, we need two sets of data: a big training set, and a smaller validation set. The network is trained on the training set, and tested on the validation set, so it is very important that there are no images that are in both sets.
./train.py --width 128 --height 64 --length 4 --symbols symbols.txt --batch-size 4 --epochs 2 --output-model test.h5 --train-dataset training_data --validate-dataset validation_data
Train the neural network for 2 epochs on the data specified. One epoch is one pass through the full dataset.
The suggested training dataset size for the initial training for captcha length of 4 symbols is 20000 images, with a validation dataset size of 4000 images.
./classify.py --model-name test --captcha-dir ~/Downloads/validation_data/ --output ~/Downloads/stuff.txt --symbols symbols.txt
With --model-name test the classifier script will look for a model called
test.json with weights test.h5 in the current directory, and load the model
up.
The classifier runs all the images in --captcha-dir through the model, and
saves the file names and the model's guess at captcha contained in the image in
the --output file.