In this thesis, an automatic products recognition system has been developed using deep learning techniques. After a detailed study of the state-of-the-art object
detection frameworks and the image classification networks, the dataset has been
acquired and manually labelled. Several frameworks have been trained and tested to
select the most suitable for this application. Results of each model are provided in
this thesis, together with some considerations. At the end, the most performing object detection system has been encapsulated inside a web service environment that
provides a friendly and easy-to-use interface to load an image, detect the products
and return the results. The entire project has been developed using Python with
TensorFlow as deep learning framework and Flask as web service environment.