52 lines
2.0 KiB
Python
52 lines
2.0 KiB
Python
"""
|
|
Usage:
|
|
python scripts/annotate.py /home/iriley/Documents/pdf/scanned/10.pdf 5 --type table --show
|
|
python scripts/annotate.py /home/iriley/Documents/pdf/scanned/10.pdf 5 --type redaction --show
|
|
python scripts/annotate.py /home/iriley/Documents/pdf/scanned/10.pdf 5 --type layout --show
|
|
python scripts/annotate.py /home/iriley/Documents/pdf/scanned/10.pdf 5 --type figure --show
|
|
"""
|
|
|
|
import argparse
|
|
|
|
from cv_analysis.utils.display import show_image
|
|
from cv_analysis.utils.draw import draw_contours, draw_rectangles
|
|
from cv_analysis.utils.open_pdf import open_pdf
|
|
from cv_analysis.utils.visual_logging import vizlogger
|
|
|
|
|
|
def parse_args():
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("pdf_path")
|
|
parser.add_argument("--page_index", type=int, default=0)
|
|
parser.add_argument("--type", choices=["table", "redaction", "layout", "figure"], default="table")
|
|
parser.add_argument("--show", action="store_true", default=False)
|
|
args = parser.parse_args()
|
|
return args
|
|
|
|
|
|
def annotate_page(page_image, analysis_function, drawing_function, name="tmp.png", show=True):
|
|
result = analysis_function(page_image)
|
|
page_image = drawing_function(page_image, result)
|
|
vizlogger.debug(page_image, name)
|
|
show_image(page_image)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
args = parse_args()
|
|
page = open_pdf(args.pdf_path, first_page=args.page_index, last_page=args.page_index)[0]
|
|
name = f"{args.type}_final_result.png"
|
|
draw = draw_rectangles
|
|
if args.type == "table":
|
|
from cv_analysis.table_parsing import parse_tables as analyze
|
|
elif args.type == "redaction":
|
|
from cv_analysis.redaction_detection import find_redactions as analyze
|
|
|
|
draw = draw_contours
|
|
elif args.type == "layout":
|
|
from cv_analysis.layout_parsing import parse_layout as analyze
|
|
elif args.type == "figure":
|
|
from cv_analysis.figure_detection.figure_detection import detect_figures
|
|
|
|
analyze = detect_figures
|
|
annotate_page(page, analyze, draw, name=name, show=args.show)
|