Sensovision Systems has developed an classficaiton model using the ResNet50 network architecture to classify surface of a metal parts using machine vision cameras. This model can trained on a dataset of 10-20 images, and the output is the classification of parts as top or bottom layer. This project will help in better part orientation with higer accuracy Here’s an example of how you can use the model to Classifiy parts:

1. Capture an image using a camera.

2. Feed the image to the model.

3. Get the type of output for each image.