chessboard-image-to-fen

Takes an overhead picture of a chessboard and converts it to FEN notation for engine analysis

Running server

screen -S chess_flask

screen -x -R chess_flask

cd ~/chessboard-image-to-fen mkdir saved_objects conda activate tf-cpu psudo() { sudo env PATH="$PATH" "$@"; } git pull psudo python3 app.py

screen -x -R file_checker cd ~/neural-chessboard/ conda activate tf-cpu git pull python3 file_checker.py

steps to annotate new training data:

take photos put photos in neural-chessboard/chess_1857 and edit detect_mass.sh run ./detect_mass.sh delete all malformed chess boards in output folder move files that exist in chessboard-image-to-fen/clayboards_out/ to chessboard-image-to-fen/clayboards_out_processed/ move files from neural-chessboard/clayboards_out output folder to chessboard-image-to-fen/clayboards_out/ label all images in labeled_boards.py run create_training_data_from_boards.py ensure new images exist in training folders and validate images there to ensure they're good move files that exist in chessboard-image-to-fen/clayboards_out/ to chessboard-image-to-fen/clayboards_out_processed/

commands to setup ec2 instance:

sudo yum install git -y y

cd ~

git clone git@github.com:cschubiner/chessboard-image-to-fen.git mkdir saved_objects

git clone git@github.com:cschubiner/neural-chessboard.git yes

conda create -n tf-cpu tensorflow y

conda activate tf-cpu y

cd neural-chessboard

pip3 install --no-cache-dir -U git+https://github.com/chsasank/image_features.git

pip3 install --no-cache-dir -r requirements.txt

pip3 install --no-cache-dir keras

pip3 install --no-cache-dir pandas

pip3 install --no-cache-dir flask

python3 -m pip install --no-cache-dir -U git+https://github.com/chsasank/image_features.git

python3 -m pip install --no-cache-dir -r requirements.txt

python3 -m pip install --no-cache-dir keras

python3 -m pip install --no-cache-dir pandas

python3 -m pip install --no-cache-dir flask

conda install opencv y

conda install keras y

conda install pyclipper y

conda install numpy y

conda install scipy y

python3 -m pip install pyclipper

conda install matplotlib y

conda install --file requirements.txt y

conda install scikit-learn y

conda install pandas y

pip3 install --no-cache-dir -U git+https://github.com/chsasank/image_features.git

python3 -m pip install --no-cache-dir -U git+https://github.com/chsasank/image_features.git

cd ~/neural-chessboard python3 main.py detect --input="clayboards/IMG_1353.jpg" --output="clayboards_out/board_1353.jpg"

cd ~/chessboard-image-to-fen/ python3 eval_classify.py

Google Cloud AutoML Vision

make sure you run these commands in the google cloud shell of the project that has autoML in it

PROJECT=$(gcloud config get-value project) && BUCKET="${PROJECT}-vcm" gsutil mb -p ${PROJECT} -c regional -l us-central1 gs://${BUCKET}

gcloud projects add-iam-policy-binding $PROJECT
--member="serviceAccount:custom-vision@appspot.gserviceaccount.com"
--role="roles/ml.admin" gcloud projects add-iam-policy-binding $PROJECT
--member="serviceAccount:custom-vision@appspot.gserviceaccount.com"
--role="roles/storage.admin" gcloud projects add-iam-policy-binding $PROJECT
--member="serviceAccount:custom-vision@appspot.gserviceaccount.com"
--role="roles/serviceusage.serviceUsageAdmin"

Notes

If you get this error with image_features, follow steps here: https://stackoverflow.com/questions/50236117/scraping-ssl-certificate-verify-failed-error-for-http-en-wikipedia-org

Links

https://chsasank.github.io/deep-learning-image-features.html https://github.com/chsasank/image_features

cschubiner / chessboard-image-to-fen