Development of a Gemstone Type Identification System Based on HSV Space Colour Using an Artificial Neural Network Back Propagation Algorithm

Abstract

A gemstone is a mineral stone that be formed from the result of geological processes and has a hardness above 7 Mohs. Nowadays, gemstones have become famous in Indonesian society. Many facts concerning the business of gemstone rings, including encouragement by the the central government for the gemstone souvenirs that are given to State guests by the President of Indonesia, gemstone contests and gemstone exhibitions have all contributed to pique the interest of researchers in the subject. This current research is undertaken to create a system that is able to identify three kinds of gemstone, namely, Ruby, Sapphire and Emerald using the Hue Saturation Value (HSV) colour space, image processing techniques and an Artificial Neural Network (ANN) back propagation algorithm by learning from examples. The hue from the HSV colour space of a gemstone will be used by the system for training. In the image processing, the system will crop, resize, convert from RGB to HSV, obtain a Hue colour and extract the colour as a 30x3 matrix. The extracted result will be used to train an ANN consisting of three input layers, three hidden layers and one output layer with targets that have been pre-determined. The results of the tests showed a degree of accuracy of 90.66% with 5 times of training and 25 times of testing on any type of gemstone. The result shows that using Artificial Neural Network Back Propagation in identification gemstone types is success, because the accuracy system has a highest percentage.

Automatic Gemstone Classification Using Computer Vision

“…Stacking Support Vector Machine, Logistic Regression and Multilayer Perceptron models further improved the accuracy by 0.3%. Despite the robustness of computer vision systems in mineral recognition, only one study on automatic identification of gemstone images [30] has been reported to date, to the best of the authors' knowledge. A per-class accuracy of 75-100% was attained for classifying unseen Ruby, Blue Sapphire and Emerald images based on the Hue channel of the HSV colour space using an Artificial Neural Network.…”