Bench-Top Plastic Contaminant Detection in Cotton Using Deep Learning Neural Network Trained with Images Taken Under 4 Lighting Colors

Joshua H Tandio , Mississippi State University
Filip To , Mississippi State University
Ruixiu Sui , USDA-ARS
It is about detecting plastic contaminant in cotton using camera and deep learning neural network. The camera will capture image of cotton and the neural network will classify the image as containing plastic contaminant or not containing plastic contaminant. The neural network model is trained with training images to detect plastic contaminant, and we also study the effect of lighting color used for taking these training images in the model's detection accuracy.