| In recent years,with the popularity of consumer drones,the number of civilian drones is gradually increasing.However,due to the imperfect supervision mechanism and relevant laws and regulations,"black flying" UAV incidents occur frequently,posing a serious threat to social order and public safety.In the urban environment,the existing detection methods have problems of false detection and missing detection of non-metallic UAV with low flight height and small size.In this paper,the spectral imaging detection technology is introduced.From the perspective of spectral characteristics of materials,the spectral characteristics of UAV target materials in urban environment are studied,and the optical detection of UAV targets is realized based on the spectral information.First,in the visible band of 350-1000 nm,spectral characteristics difference between UAV target materials and common target materials in urban environment is studied based on hyperspectral imaging means,and effective imaging feature bands of UAV targets in urban environment are extracted.Then,in 8-12μm long-wave infrared band,the finite element analysis method is used to simulate and analyze the infrared radiation characteristics of UAV,and the infrared radiation characteristics of UAV are obtained.In view of the problems that the temperature of UAV is close to the ambient temperature in the daytime and the features of UAV are not obvious in infrared images,the fusion method of visible and infrared images is studied by combining the characteristics of visible and infrared wavelengths of UAV targets,so as to enhance the contrast difference between targets and the background and provide theoretical method support for UAV target detection in urban environment.The main research contents of this paper are as follows:(1)Three different UAV materials were selected for molecular composition determination.Qualitative analysis of these three materials was carried out through Fourier infrared spectroscopy,and the main molecular components of the three UAV materials were measured as ABS plastic,PC plastic and carbon fiber reinforced epoxy resin,which laid the foundation for the subsequent study of spectral characteristics.(2)Study the selection of effective imaging feature bands of UAV targets in urban environment.Hyperspectral image data of UAV and background in urban environment were collected by hyperspectral imaging equipment,spectral database was constructed,and data processing and analysis were carried out.The feature band selection criterion of target imaging is studied based on the difference of target and background optical reflection intensity in different bands.The feature band selection method based on spectral reflectance,image information content and image brightness characteristics is given,and the imaging band suitable for UAV target detection in the range of 350-1000 nm is obtained.(3)The radiation characteristics of UAV target infrared thermal imaging were simulated and verified by experiments.COMSOL simulation software was used to simulate the infrared radiation characteristics of UAV targets,and the temperature distribution law was analyzed during the UAV flight.The results show that the temperature distribution on the body surface of the UAV is uneven,and the temperature on the upper surface of the battery and the motor is high.The reliability of the simulation results is verified by field experiments.(4)Carry out multi-spectral fusion based on feature band visible image and infrared thermal radiation image.Firstly,the image enhancement algorithm is studied to improve the quality of UAV infrared image,so that the UAV target features are more obvious.Then,the multi-spectral image fusion algorithm of UAV target imaging feature band image and infrared image is studied.Field experiment results show that the fusion image can effectively improve the feature of the target and suppress the background,effectively enhance the difference between the target and the background,and help to eliminate false alarms. |
