Security-Application-using-Image-Classification

The aim of this project is to increase automation in security applications by classifying the incoming object as permissible or not permissible. For illustration purposes, the permissible object folder consists of toddler images sourced from Google and the not permissible object folder consists of pet images sourced from Kaggle. This project uses Keras backend and creates a Convolutional Neural Network. The result obtained was 93.47% accuracy.