Research on Target Location Algorithm Based on UAV Monocular Vision

Aiming at the problem of UAV space target positioning, this paper proposes a monocular visual multipoint locating method based on photogrammetry technology and UAV inertial navigation information. The paper establishes the three-dimensional coordinate error equation of obstacles based on the principle of three points collinear. In the process of rigorously solving the error equation, the paper uses a method based on Tikhonov regularization and L-curve to reduce the noise interference in the positioning process and meet the positioning accuracy requirements. Finally, this paper builds a hardware platform for UAV positioning to verify this method, and compares it with the least squares method. It is verified that this method is simple and easy to implement, accurate positioning, and strong anti-interference ability.