update:Add md make mode config in do_parse.You can control whether the produced md is for NLP or MM by changing the value of f_make_md_mode

# Conflicts:
#	magic_pdf/model/pdf_extract_kit.py

1.Disable scaling when loading large images.
2.Moving the logic for channel conversion in image processing.

fix: The presence of ".pdf" multiple times in the pdf_path results in model_path not matching the expected.