The Classification Method of Urban Architectural Styles Based on Deep Learning and Street View Imagery

Abstract

The task of identifying urban architectural styles occupies a very necessary position in the fields of construction of smart cities, sustainable urban development and community regeneration. The research method proposed in this paper can improve on the inconveniences of traditional methods of identifying urban architectural styles, such as: the community building is relatively old, and the integration of more periods of architectural style can significantly affect the test results. It is an established fact that data cannot be collected and processed efficiently by humans alone, and can not enter such qualitative and descriptive research methods into the computer for auxiliary research. This paper is based on the explosion of information data use in the 21st century, and use deep learning technology to process unstructured data with convolutional neural networks as the core to assist in the identification of urban architectural styles. With the rapid development of deep learning technology in recent years, its classification techniques for identification of street images of urban buildings can be used for urban management, and a new strong underpinning for the allocation of urban resources, urban diversification management, and the transformation of old communities in the later period has been provided by the proper classification of urban architectural styles. Notwithstanding its restrictions, the approach presented in this research has shown promise and the valuable value of deep learning-based techniques for the study of architectural styles, and this approach has universal significance.

Urban Architectural Style Recognition and Dataset Construction Method under Deep Learning of Street View Images: A Case Study of Wuhan

“…However, some studies have limited application and are unsuitable for large-scale urban applications. The author proposes the use of deep learning technology based on CNNs combined with Wuhan SVIs to identify and verify the dominant style of Wuhan architecture [50,51]. Taking Wuhan architecture as the research object, the aim of this study is to identify and classify Wuhan's architectural style and construct a Wuhan architectural style dataset, demonstrating the viability of the suggested approach for widespread architectural style recognition in cities.…”

A Spatial Analysis of Urban Streets under Deep Learning Based on Street View Imagery: Quantifying Perceptual and Elemental Perceptual Relationships

“…In recent years, studies on street space evaluation have become more prosperous and more diversified in terms of methodology [14,15]. Traditional research on street spatial quality is mainly carried out using manual measurements and subjective judgment, which brings about specific difficulties for refined street spatial evaluation.…”