Skip to content

Latest commit

 

History

History
17 lines (13 loc) · 903 Bytes

README.md

File metadata and controls

17 lines (13 loc) · 903 Bytes
Raw

Semi-Supervised-Bounding-Box-Predictor

An intelligent tool to predict fitting polygons given the bounding box around the region.

This system takes in an input image and Bounding Box coordinates around a region for annotation in document image and outputs a tight fitting bounding polygon around the region.

Install prerequisites

Install requirements stated in requirements.txt

To run Inference on your own image

  1. Download the model from this link
  2. Place the Final.pth file in the folder (Semi-Supervised-Bounding-Box-Predictor/ssbbox/)
  3. Run the script(Semi-Supervised-Bounding-Box-Predictor/ssbbox/toolscript.py) as ->
python3 toolscript.py --img path/to/your/image --bbox x,y,w,h
  1. This outputs the final result in terms of X and Y (in new lines) coordinates respectively.