Analysis and implementation of an application for analysis of paintings Pinoteka

Abstract

Pinoteka is an agile solution that allows the user to access demanding arts through built-in innovative technologies. The program is a dynamic web application created based on a client-server model. On the client side, it is supported by React.js a JavaScript library and on the server side, it is utilized with Python programming language and Django web framework with a database supported by Azure SQL Service and Database with Azure Blob Storage. It uses deep learning-based image processing algorithms to match the given image input with the existing image. Each match will provide information about that given image. Our strategy for handling this process is via image models stored in the cloud and vectors kept in databases. It will not only improve accuracy but also help with the agility of our program. The paper contains a detailed description of image recognition and comparison methods. It compares the research of the topic and the web application containing image features comparison experiment. It explains the necessary topics to cover in advance and the comparison of them. Demonstrates non-technical individuals’ image comparison concepts and exposes to artificial intelligence. Furthermore, contains the description of the approaches and the usage of the tools and methods.