Implementation of OpenCV and CNN-Keras in Detecting Plastic Waste

Plastic waste has become a worldwide concern, as it can create a bad environment. To reduce it, better waste management is needed. The simplest waste management is waste sorting. However, the lack of understanding of trash types makes waste management awful. Thus, instead of doing waste sorting manually, it is better to have technology to help with waste sorting. Based on that, this research aims to implement computer vision and CNN to optimize waste sorting of plastic waste from other kinds of waste. The result of this research shows the model is quite good for waste sorting and has an accuracy of 75.08% for the testing dataset. However, in the implementation of the CNN model, it is more considered to be used considering the computational and deployment costs in the wider society.