This paper investigates a monocular vision-based method for Unmanned Aerial Vehicles (UAVs) target recognition and tracking, addressing the challenges of decreased recognition accuracy or tracking failure due to external interference when UAVs are equipped with visual sensors for target recognition and tracking. Design an edge detection algorithm based on adaptive Canny, considering disturbances such as changes in lighting and viewing angles. Utilize dynamic thresholding to extract edge features of targets under various environmental interferences. An adaptive template matching method based on image pyramid strategy is proposed to generate multi-angle and multi-scale image templates, which are matched with the target image to achieve target recognition. To address the issue of target tracking loss in dynamic scenes, an improved Kernel Correlation Filter (KCF) algorithm is employed based on the results of target recognition. This algorithm tracks detected stationary or moving targets by assessing target confidence, thus enhancing tracking accuracy while conserving computational resources. Flight test experiments verify that the Quadrotor UAV equipped with a monocular camera can effectively identify stationary or moving targets on the ground and achieve stable target tracking.