High-altitude unmanned aerial vehicle landing mark identification method based on cascade classifier

Abstract

The invention relates to an unmanned aerial vehicle autonomous landing guidance technology, provides an unmanned aerial vehicle landing guidance method which is slightly influenced by height, and provides a cascade classifier-based high-altitude unmanned aerial vehicle landing mark identification method, which comprises the following steps of: 1) controlling an unmanned aerial vehicle to be near a landing point; 2) acquiring an image: acquiring the image; 3) using a cascade classifier based on Haar Harr features to carry out landing mark detection on the image; (4) judging whether a landing mark exists or not, and if not, carrying out the step (5); if so, carrying out step 8); 5) performing landing mark detection on the image by using a cascade classifier based on a local binary pattern (LBP) feature; (6) judging whether a landing mark exists or not, and if not, carrying out the step (7); if so, carrying out step 8); 7) adjusting the attitude of the unmanned aerial vehicle and performing the step 2); and 8) controlling the descending height of the unmanned aerial vehicle for landing. The method is mainly applied to unmanned aerial vehicle autonomous landing occasions.

Description

technical field [0001] The invention relates to the autonomous landing guidance technology of unmanned aerial vehicles, in particular to a method for recognizing landing signs of high-altitude unmanned aerial vehicles based on cascade classifiers. Background technique [0002] Since the attitude and speed of the UAV are required to be stable within a certain range when landing, the machine vision algorithm is applied during automatic landing, and the specific landing image signs captured during the landing process are processed to obtain the attitude and orientation of the UAV, so as to guide the unmanned aerial vehicle. The man and the machine landed safely. Compared with the traditional landing method, the method based on machine vision has the advantages of high precision, low cost and strong anti-interference ability. [0003] The commonly used vision-based landing guidance technology uses the principle based on mark detection, and the mark to be detected generally adop...

AI Technical Summary

Problems solved by technology

[0004] The traditional landing guidance technology based on sign detection is affected by the height of the drone, and sometimes it cannot accurately identify the landing sign. Due to the random landing site terrain and environmental factors, the existing navigation technology can only guide the drone to the landing site. At a certain height, the inability to accurately identify the landing sign will affect the flight efficiency of the UAV, and threaten flight safety in severe cases

Images