Abstract:
"gemstone is a mineral stone with a hardness of greater than 7 Mohs due to geological activity. Gemstones have achieved significant market share in Sri Lankan society. In the current day and age, gemstones have impacted the jewellery industry's cultural beliefs. The value of gemstones is increasing, and their demand is rising. It has arrived at a point where automating gemstone identification is essential. There is no proper automated system to identify gemstones and predict price with less time and high accuracy than manual valuations. It can be a significant problem for gem buyers and sellers as their identification abilities depend on years of experience. This is a significant flaw in the gemstone industry, as substantial growth has shown during the past few years.
Using image processing techniques, Convolutional Neural Network (CNN) algorithms, and image segmentation methods, this project aims to develop a system that can recognize and determine the pricing of gemstones such as Blue Sapphire, Yellow Sapphire, Pink Sapphire, and Alexandrite. This system involves capturing a photo of the gemstone to classify and generate calculations to estimate its market price.
The Gem Identification and Price Prediction study is the first phase of an ongoing research project, and it presents a unique perspective on Sri Lanka's gem industry. The result generated from the system aims to produce good values; however, it can further be researched and developed."
