The script inserts watermark image into each image existing in input directory. Watermarks are inserted at suitable areas of an image. By default image areas are considered suitable if they have no sharp transitions (edges). I.e. consider_edges execution argument is set to 1 by default. OCR (Tesseract-OCR) and dark pixels presence check can be turned on to not allocate watermarks if there are detected text or dark pixels (in amount of greater than DARK_PXLS_ALLOWED_AMOUNT literal) in an image area respectively. Therefore, it is not recommended to turn on the dark pixels presence check for images with black or dark background. To turn on Tesseract-OCR, specify --use_tesseract 1 script execution argument. To turn on dark pixels presence check, specify --consider_dark_pixels 1 script execution argument. Use of Tesseract-OCR increases script execution time but can reduce amount of watermark wrong allocations, especially when text color doesn't much contrast with image background.
There are some script execution arguments. Use -h execution argument to see their description or look at main.py itself. Watermark image is checked for existence. In the cases when watermark image doesn't exist, input directory doesn't exist or input directory has no images, the script prints an error message and stops.
used_areas_percentage execution argument specifies how many watermarks are inserted onto all suitable areas. Default coverage is 70. In the case when used_areas_percentage or amount of founded suitable areas are too low, and the amount of watermarks to insert is 0, then 1 watermark is inserted (if there is at least 1 suitable area). When there are no suitable areas at all for an input image, corresponding warning is printed.
Demo with default execution arguments:
Inverted demo with default execution arguments (pay attention that consider_dark_pixels script execution argument should be not specified or be set to 0 for such images having dark background and light font color):
Demo with --min_scale 0.1 --used_areas_percentage 100:
Used for photo with default execution arguments (https://www.publicdomainpictures.net/en/view-image.php?image=305171&picture=yarrow-blossom):
Used for photo with used_areas_percentage script execution argument set to 15 (https://www.publicdomainpictures.net/en/view-image.php?image=305171&picture=yarrow-blossom):
Photo processed with used_areas_percentage script execution argument is set to 100 and no suitability check is turned on (--consider_edges 0 --consider_dark_pixels 0) (https://www.publicdomainpictures.net/en/view-image.php?image=305171&picture=yarrow-blossom):
The following packages are used:
- argparse 1.4.0 (by some reason it doesn't show up when using
pip freeze); - numpy 1.19.4;
- opencv-python 4.4.0.46;
- Pillow 8.0.1;
- progress 1.5;
- pytesseract 0.3.6.
The script was tested on Windows 10 64-b.