DRONE POSITIONING SYSTEMS THAT USE DIGITAL CAMERAS

Abstract

Unmanned aerial vehicles are very important in everyday life. Their number is increasing every day, as well as the scope of their use. Therefore, it becomes necessary to automate their flight from departure to kindergarten. During automatic flight, drones have certain problems during takeoff, landing and positioning. The problem is that when landing and taking off, it is necessary to ensure high positioning accuracy, which is impossible when using GPS, as it can provide accuracy of only a few meters, which is not enough, and the use of operators is accurate, but this method requires the use of quality cameras with stabilization. These stabilization cameras are very heavy, so the payload of drones is reduced, and they are very expensive (usually more expensive than the drone itself). Also, the use of operators during landing and departure can lead to a catastrophe due to human factors. The task of this article is to create a classification table, analyze landing methods, assess their advantages and disadvantages, give recommendations for the use of the most effective positioning system, as well as the development of new positioning methods.
 In the course of work modern, and also the most widespread methods of positioning were considered, the critical analysis of robots is made. As a result, it was proposed to classify drone positioning systems that provide reliable takeoff, landing and delivery of goods using digital cameras. This classification includes all combinations of digital cameras and radiation sources that can be located both on the drone and on the landing or cargo delivery area. Examples are given for each combination proposed in the classification. A thorough analysis of the advantages and disadvantages of each configuration of digital cameras and radiation sources is given. Recommendations for choosing the best drone positioning system are provided. The main disadvantages of these systems are the complexity of algorithms, which makes systems more expensive, as well as complicates the creation of the system, which does not preclude the possibility of making a mistake when creating a system. And this can lead to an accident. All DPSs can be classified on group depending on the number of digital cameras, number and shape of reference light sources, locations of digital cameras and location of the light sources. From the point of reliability and economy the best DPS should include one camera on a drone and a minimal set of reference light sources on the ground. The authors suppose that three reference light sources that specify a triangle is the best choice because it makes possible estimation of the distance and angular coordinates of the landing pad.