This Python program renames GIF files based on the text extracted from the GIF using the Tesseract OCR engine. The program categorizes the renamed GIFs into three folders based on the accuracy of the extracted text:
- High Accuracy (90-100%)
- Medium Accuracy (50-90%)
- Low Accuracy (<50%) or Error
Each renamed GIF file will have a corresponding entry in a CSV file stored in the same directory. The CSVs will contain both the original filename and the new filename for reference.
This program was created as part of the requirement to rename a large collection of GIFs based on their content, with specific focus on GIFs needed for the "Sign with Robert" project. This project can be found on GitHub: two-way-sign.
- Rename GIF files based on text content extracted using Tesseract OCR.
- Sort GIFs into directories based on accuracy thresholds.
- Avoids duplicating words in filenames.
- Logs original and new filenames in CSV files.
- Skips files that are in use by another process (i.e., file locked by another application).
- Option to enable/disable debug mode for detailed logging.
The following dependencies are required to run the program:
- Python 3.x
- Tesseract OCR
- Python Libraries:
pytesseractPillowshutilcsvostkinter
Make sure you have Python 3.x installed on your system. You can download it from the official website.
Tesseract OCR must be installed and configured to work with the program. You can download it from here.
Once installed, configure the pytesseract.pytesseract.tesseract_cmd to point to the Tesseract executable.
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'