98 lines
2.5 KiB
Bash
Executable File
98 lines
2.5 KiB
Bash
Executable File
#!/bin/bash
|
|
|
|
echo "#"
|
|
echo "# phase 1: take screenshots"
|
|
echo "#"
|
|
|
|
if [ ! -d "screenshots" ]; then
|
|
mkdir screenshots
|
|
fi
|
|
|
|
find . -iname "*.mp4" -print0 | while read -d $'\0' file
|
|
do
|
|
filename=$(basename -- "$file")
|
|
#extension="${filename##*.}"
|
|
filename="${filename%.*}"
|
|
ffmpeg -ss 0 -i $file -vf fps=0.5 ./screenshots/$filename-%d.png
|
|
done
|
|
|
|
echo "#"
|
|
echo "# phase 2: img preprocess"
|
|
echo "#"
|
|
|
|
if [ ! -d "preproccessed" ]; then
|
|
mkdir preproccessed
|
|
fi
|
|
find "screenshots" -iname "*.png" -print0 | while read -d $'\0' file
|
|
do
|
|
filename=$(basename -- "$file")
|
|
extension="${filename##*.}"
|
|
filename="${filename%.*}"
|
|
convert "$file" -gravity SouthWest -crop 560x48+0+8 -fuzz 12% +transparent "#edf1c9" "preproccessed/$filename.png"
|
|
done
|
|
|
|
echo "#"
|
|
echo "# phase 3: ocr"
|
|
echo "#"
|
|
|
|
if [ ! -d "ocr" ]; then
|
|
mkdir ocr
|
|
fi
|
|
find "preproccessed" -iname "*.png" -print0 | while read -d $'\0' file
|
|
do
|
|
filename=$(basename -- "$file")
|
|
extension="${filename##*.}"
|
|
filename="${filename%.*}"
|
|
tesseract $file "ocr/$filename" -c tessedit_char_whitelist=" 0123456789KM/HNW."
|
|
sed -i 's/ //g' "ocr/$filename.txt"
|
|
|
|
done
|
|
|
|
echo "#"
|
|
echo "# phase 4: prep for geotagging"
|
|
echo "#"
|
|
if [ ! -d "tagged" ]; then
|
|
mkdir tagged
|
|
fi
|
|
find "screenshots" -iname "*.png" -print0 | while read -d $'\0' file
|
|
do
|
|
filename=$(basename -- "$file")
|
|
extension="${filename##*.}"
|
|
filename="${filename%.*}"
|
|
convert $file "tagged/$filename.jpg"
|
|
done
|
|
exiftool -overwrite_original -TagsFromFile ./knowngood.jpg -All:All "tagged/*.jpg"
|
|
|
|
echo "#"
|
|
echo "# phase 5: if ocr was good, geotag. Otherwise, reject."
|
|
echo "#"
|
|
goodreads=0
|
|
badreads=0
|
|
if [ ! -d "badread" ]; then
|
|
mkdir badread
|
|
fi
|
|
find "ocr" -iname "*.txt" -print0 | while read -d $'\0' file
|
|
do
|
|
filename=$(basename -- "$file")
|
|
extension="${filename##*.}"
|
|
filename="${filename%.*}"
|
|
|
|
n=$(grep -Poh "(?<=N)[0-9]{2}\\.[0-9]+" "$file")
|
|
w=$(grep -Poh "(?<=W)[0-9]{2}\\.[0-9]+" "$file")
|
|
|
|
if [ -n "$n" ] && [ -n "$w" ]; then
|
|
goodreads=$(($goodreads+1))
|
|
echo "$filename was OCR'd well; $goodreads good so far"
|
|
exiftool -overwrite_original -gpslatitude="$n" -gpslongitude="-$w" -GPSLatitudeRef="North" -GPSLongitudeRef="West" "tagged/$filename.jpg"
|
|
else
|
|
badreads=$(($badreads+1))
|
|
echo "$filename is not perfect; $badreads bad so far"
|
|
mv "tagged/$filename.jpg" "badread/$filename.jpg"
|
|
mv "ocr/$filename.txt" "badread/$filename.txt"
|
|
mv "preproccessed/$filename.png" "badread/$filename.png"
|
|
fi
|
|
done
|
|
|
|
echo "$goodreads good, $badreads bad. accuracy:"
|
|
bc <<< "scale=4; goodreads=$goodreads; badreads=$badreads; ratio=100*goodreads/(goodreads+badreads);ratio"
|