Applications of artificial intelligence recognition technology on construction industry

Abstract

In the construction industry, technology-based Artificial Intelligence (AI) has become the center of development. AI image recognition technology has been developed and applied to face various application scenarios on the construction site and help workers increase their efficiency and accuracy in recognized work. The purpose of this paper is to introduce two mature AI technologies: the reinforcement quantity identification system and the intelligent identification and evaluation method of typical damages to a fair-faced wall. The reinforcement quantity identification system uses the method to train a model on the Yolo v3 algorithm. It is able to help workers use an image taken to get the number of bars to be counted. The intelligent identification and evaluation method to a fair-faced wall is designed to focus on the three typical damages, weathering, efflorescence, and plant coverage. In the design model, the used algorithm is Yolo v4, and OpenCV is applied as a vision library to access the degree of damages. Workers only need to upload an image to the system, and then it will return an analysis result for the wall. Through this paper, readers will understand their main algorithms and models based on AI technology, as well as their practical applications in the construction industry, and provide a theoretical introduction for future AI image technology references.