Research On Image Processing Technology In UAV Landing

So far,the UAV have been widely applied,either to military or civilian.During the application of UAV,its positioning is the most vital part.However,many manufacturers now use GPS to locate it.Although this method is efficient,it is susceptible to be interfered and often has no signal in closed environments such as indoors or in tunnels,which can not meet clients’ requirements for positioning accuracy and application range.Therefore,it’s necessary to achieve the accurate positioning of UAV and break the limitations of GPS which can not properly locate in some special circumstances.It not only satisfies strict requirements of many applications for landing,but also has practical significance for the research and development of UAVs.The subject take advantage of image processing to achieve the location of four-rotor aircraft.Firstly,the gradation is operated on the collected landing mark image;the histogram of the grayscale image is equalized to enhance the contrast of the image;the image which has been subjected to the equalization operation is smoothed to filter the noise;then,several commonly used target segmentation methods are analyzed,including iterative thresholding method,basic adaptive threshold method and Otsu threshold method.And combine the subject’s physical truth,the multi-sub image segmentation method which is better in effectiveness and extraction effect is selected.In order to achieve the outline description,select the wavelet transform operator to detect the extracted edge of the mark.In order to complete the feature point extraction of the identification image,the Hough line detection algorithm is adopted.Finally,the relative position information of the aircraft and the landing mark is estimated by using the flight parameters of the rotorcraft collected by the sensor.Experiments are conducted to verify the effect of the visual assist system created by this subject on the positioning process of the UAV.The data obtained are analyzed to find the potential influencing factors.Based on digital image processing technology,this paper constructs a low-cost,simple and practical help-drop system.Through a series of processing and analysis of the collected images,the relative positions of the UAV and landing marker are calculated.At the theoretical level,it provides a certain reference.