Created: October 8, 2021 2:47 PM Last Edited Time: October 8, 2021 2:55 PM

This project is built on top of tensorflow 2 object detection api.

The training is done on free version of google colab.

Dataset is taken from kaggle and google images.

The following project is based on the defects found on the fabric at the time of its inspection. The system can detect those defects through giving the path of the target folder or through webcam real time detection or by connecting it to external camera source.

Note that the real time version needs some powerful machine to achieve FPS and external webcam will introduce lag in the detection mechanism.

The classes for defect are:

1. Hole
2. Stain