Unmanned Autonomous Helicopter Landing Algorithm And Implementation

Nowadays , artificial intelligence is the trend of science and technology. Syncretizing the strongpoints of artificial intelligence and unpiloted helicopter, the unpiloted helicopter can accomplish the assignments that the manned helicopter can't. So the unpiloted helicopter has enormous application potential. So the primary great nations have mustered relevant human resource and financial support in this region for researching robot helicopters. Using image processing techniques on unmanned helicopter is very mainstream.Thisarticle referred to prevenient research findings, introduced advantages of unmanned helicopter, discussed its application and study progress, had a close study of its landing and optical system principle. The article has consulted previous successful efforts and improved visual algorithms and increased Operational Speed. This article use the new approach that two cameras is installed under the bottom of the helicopter with two primary optical axis parallel to each other, So the cameras may be applied to unary operation and binocular vision. When using unary operation, one camera works and when using binocular vision, two cameras work.This paper covers structurally two parts. The first part is unary operation and the second part involves binocular vision. When the helicopter lands on warships Unary operation is used and binocular vision is used when the helicopter lands on complex landform. There's two unary algorithm Programs because the first one is not very successful. In binocular vision program, There're two reasons we use the method that two cameras'primary optical axis are parallel to each other, firstly, we can simplify epipolar constraint arithmetic so that arithmetic speed and precision can be increased. Secondly, the unary operation should be compatible with existing hardware. Because two cameras'primary optical axises are perpendicular to the bottom of the helicopter body, when one camera is started, we can use unary operation algorithm.The paper did many works to realize the algorithm of unary vision and binocular vision. On the part of unary vision, the icon and algorithm in the second Proposal is very simple and easy to deal with. They have the advantage of quick and accurate. The paper also set up the mathematical model. it used rotation matrix and translation matrix to establishing the functional relation between helicopter coordinate system and the deck coordinate system, and get helicopter's position and posture by solving linear equations with one unknown. On the part of binocular vision, the article first developed mathematical models. For the difficulty and time-consuming stereo matching, the paper expounded the comparative experimental investigation on traditional matching method and modified matching method. For the critical matching constraint condition--- epipolar constraint, the paper suggested a more simple and efficient and fast constraint condition--- parallelism of optical axis epipolar constraint, by refining the quintessence of traditional epipolar constraint, and expounded the condition.In every program we first explain theory and then explain implementation procedure and then show the experimental result and finally summarize the scheme.