In recent years, visual tracking in animals has been widely applied in various fields. To better capture the flapping posture of birds for tracking, we conduct target tracking of a bioinspired bird by using both traditional methods and deep-learning methods. We estimate the pose of the bird based on visual perception. For the tracking part, we collect and create a flight dataset of a bioinspired bird indoors and outdoors. The deep learning-based pytracking algorithms are used for tracking and evaluation. For the pose estimation part, we first collect data on the bird, selected feature points, and used the labelimg tool to annotate and obtain image coordinates. The pose estimation problem is transformed into a perspective-n-point problem (PnP). The pose is solved by using the EPnP method, successfully obtaining the pose information in the experimental photos. We have also compared it with the RPnP method. EPnP method is more convenient and efficient. This paper proposes a purely visual method for pose estimation of the bioinspired bird. We defines the posture, and provides the possibility for solving the posture. The results of the project can be applied to fields such as wild bird tracking, military defense systems, bird flight performance analysis, and unmanned aerial vehicle ground visual perception systems.