Automatic Classification of Semi Precious Rocks

Irving Alberto Cruz Matías, Constantino Pearl, Andrea Puente

Research output: Contribution to conferencePaperpeer-review


This project pretends to reach a successful model of automatic classification of semi precious rocks through Machine Learning techniques in TensorFlow and OpenCV image processing algorithms. Visual characteristics and image processing algorithms were proposed to correctly seg-ment the objects and identify their key features. After that, different TensorFlow models (DenseNet, NasNet, etc.) were tested to measure their accuracy and select the best method based on a comparative between performance and precision.
The results from the experiments were assessed and the definite algorithm was construct-ed. The algorithm runs on an Amazon Web Services instance, which is accessed by a mobile application. Results, scope and project limitations are discussed at the end of this work, as well as future approaches.
Original languageEnglish
Publication statusIn preparation - 2019
Event2020 Winter Conference on Applications of Computer Vision - Snowmass Village, Colorado, United States
Duration: 2 Mar 20205 Mar 2020


Conference2020 Winter Conference on Applications of Computer Vision
Abbreviated titleWACV20
Country/TerritoryUnited States
Internet address


Dive into the research topics of 'Automatic Classification of Semi Precious Rocks'. Together they form a unique fingerprint.

Cite this