The unknown goaf in coal mines poses a potential danger to coal mine safety production. Hence, detecting its location and shape is very important to ensure coal mine safety production. Although numerous methods are currently available for detecting goafs, they are expensive, inefficient, and have a long cycle. In recent years, with the rapid development of dense array observation technology, the ambient noise imaging technique has been widely used in near surface structure detection. In this study, we used the dense array observation based on the nodal seismometers and ambient noise imaging technology for goaf detection in the Haerwusu open-pit coal mine. We deployed 145 stations with 16 m spacing for 10 consecutive days. Our results show that: (1) The VS anomalies were well constrained using the ambient noise imaging and ESPAC methods. (2) Potential locations of the goaf were detected based on the distribution of the low VS regions. After verification by drilling, the locations showed accurate correspondence with the positions revealed by the borehole, which proves the validity of this method for goaf detection. Our study shows that the dense seismic array technique for ambient noise imaging is an economic and effective means of conducting goaf detection in open-pit mines and the technique has a wide range of applications.