Iranian Architectural Styles Recognition Using Image Processing and Deep Learning

Iranian architecture, also known as Persian architecture, encompasses the design of buildings in Iran and extends to various regions in West Asia, the Caucasus, and Central Asia. With a rich history dating back at least 5,000 BC, it boasts distinctive features and styles. Iran, located in the Middle East, has faced ongoing geopolitical challenges, including the potential for conflicts, such as those in Iraq and Afghanistan. Unfortunately, historical monuments often become unintentional casualties during wartime, suffering damage or destruction. These historical monuments hold cultural and historical significance not only for the country they belong to but for all of humanity. Therefore, it is crucial to make efforts to preserve them. In this paper, we propose the development of an automated system utilizing Deep Learning methods for the detection and recognition of historical monuments. This system can be integrated into military equipment to help identify the architectural style of a building and determine its construction date. By doing so, it can provide a critical warning to prevent the targeting of historically significant structures. To support our system, we have curated a dataset consisting of approximately 3,000 photographs showcasing six distinct styles of Iranian historical architecture. Figure 1 provides some examples of these photographs. It is worth noting that this dataset can be valuable for various scientific research projects and applications beyond our proposed system. Additionally, it offers tourists the opportunity to learn about Iranian historical monuments independently, using their mobile phones to access information about a monument’s historical period and architectural style, eliminating the need for a traditional guide. This initiative aims to safeguard the invaluable cultural heritage of Iran and neighboring regions, contributing to the collective preservation of these historical treasures.