As the number of cases where human life and property are threatened by illegal use of UAV is rapidly increasing, a technology for detecting illegal UAV using deep learning algorithm is being researched. In order for the deep learning algorithm to work, an appropriate UAV object image size and quality are required. Unfortunately, in detecting a small UAV using a camera sensor, when the object image of the camera becomes small, the detection distance of the UAV is reduced due to the difficulty of detecting the UAV. Therefore, a large number of cameras are required to expand the detection distance. This paper suggests the methodology for configuring UAV surveillance area for image acquisition in terms of reducing the required number of cameras. Four methods for constructing the UAV surveillance area considering the optical limitations of the camera are proposed, and compared the proposed methods using a simple thin lens camera model.