Image transcription with Google Images Annotation API

hand-write.sh

export GOOGLE_APPLICATION_CREDENTIALS=./path/to/google-key.json

# Settings
INPUT_IMAGE=~/Desktop/input.jpg

# Variables
DATA_PATH=~/path/to/save/transcriptions
DATE_NOW=$(date '+%y%m%d_%H%M%S')
OUTPUT_IMAGE=$(echo $DATA_PATH)/$(echo $DATE_NOW)_image.jpg
OUTPUT_IMAGE_FULL=$(echo $DATA_PATH)/$(echo $DATE_NOW)_image-original.jpg
OUTPUT_JSON=$(echo $DATA_PATH)/$(echo $DATE_NOW).json
OUTPUT_TEXT=$(echo $DATA_PATH)/$(echo $DATE_NOW)_text.md
OUTPUT_TEXT_CLEAN=$(echo $DATA_PATH)/$(echo $DATE_NOW)_text-clean.md

# Copy image
cp $INPUT_IMAGE $OUTPUT_IMAGE_FULL

# Reduce image size
convert -geometry 800x400 $INPUT_IMAGE $OUTPUT_IMAGE

# Request transcription from Google Cloud Vision API
curl -X POST \
-H "Authorization: Bearer $(gcloud auth application-default print-access-token)" \
-H "Content-Type: application/json; charset=utf-8" \
https://vision.googleapis.com/v1/images:annotate -d "{'requests': [{'features': [{ 'type': 'DOCUMENT_TEXT_DETECTION'}],'image': {'content': '$(base64 $OUTPUT_IMAGE)'}}]}" \
> $OUTPUT_JSON

# Display annotated text
TEXT=$(jq '.["responses"][0]["fullTextAnnotation"]["text"]' $OUTPUT_JSON)
echo -e "\033[1;31m $TEXT \033[0m"

# Replace line breaks with spaces to make text continuous
TEXT_CLEAN=${TEXT//\\n/ }

# Save to text file
echo -e $TEXT > $OUTPUT_TEXT
echo -e $TEXT_CLEAN > $OUTPUT_TEXT_CLEAN

# Open Markdown file
open $OUTPUT_TEXT_CLEAN