Due to the dynamic nature of cognitive radio networks, multi-user video streaming (with various video traffic characteristics and QoS requirements) requires efficient dynamic channel selection schemes to exploit available spectrum resources. To do this, a wireless user needs to effectively model the dynamic wireless environment and estimate the delay of video packet transmission when selecting a specific frequency channel. In this paper, we apply the priority virtual queue model for these wireless users to adapt their channel selection and maximize video qualities. The simulation results show that the proposed channel selection solution based on priority scheduling outperforms the conventional dynamic channel selection scheme by 2 dB (PSNR).