Along with the popularization of high-speed internet and high-performance digital cameras, the consumption prosperity of high-resolution videos and pictures appears. However, high-performance cameras bring about content infringes through illegal videotaping, which greatly damages the privileges and benefits of the content owners. Multimedia forensics is one of the most important copyright protection tools. In present forensic technologies, picture-in-picture detection is an important pre-processing step. In this study, a three-step picture-in-picture detection method is designed to effectively find the region of recaptured video. The region location is simply obtained from the I-frames after a rough detection and a fine-tuning step. Experimental results proved the effectiveness of the proposed work.