Practical interest in ‘computer vision’ has risen remarkably over the last 20 years, transforming the current state of construction-related research and attracting the worldwide attention of scholars and practitioners. This study conducts a scientometric review of the global research published between 1999 and 2019 on computer vision applications for construction, through co-author, co-citation, keyword and clustering analysis. A total of 1158 journals and conference proceedings from Scopus database were analyzed. Trends within the field are identified, as are the dominant sub-fields and their interconnections, as well as citation patterns, key publications, key research institutions, key researchers, and key journals, along with the extent to which these interact with each other in research networks. The provided results were analyzed to identify the deficiencies in current research and propose future trends. Among these is a bias in the research literature towards traditional on-site construction and a concerning gap of off-site construction research, as well as a lack of inter-relationships and collaboration between researched areas, the researchers themselves, and/or the research institutions. In the near future, computer vision will play a key role in the future development of smart construction and improvement of quality in construction projects. This study hopes to bring awareness to the industry, the journal editors, and the researchers of the need for a deeper exchange of ideas in any future research efforts.