In recent years, concrete 3D printing has advanced rapidly in the field of intelligent construction. However, factors such as printing materials, equipment, and processes make it challenging to control the forming quality of printed structures. As a non-contact detection method, computer vision technology has gradually been applied to defect detection for quality control in concrete 3D printing. Therefore, based on domestic and international research, this paper first reviews the key factors influencing the quality of concrete 3D printing, focusing on material properties and printing processes. Subsequently, it discusses traditional evaluation methods alongside those utilizing computer vision for assessing the quality of concrete 3D printing. Finally, the paper proposes a novel quality control approach that integrates deep learning and robotic arm control with computer vision technology. It provides a reference for the quality control and development of concrete 3D printing.