阅读说明：本技术 一种基于旋翼无人机的植被背景下隐藏地物的探测方法 (Detection method of hidden ground objects under vegetation background based on rotor unmanned aerial vehicle ) 是由 丑述仁 陈斌 章杰 王鹏 夏鲁瑞 苏宪程 刘甜甜 于 2019-10-24 设计创作，主要内容包括：本发明公开了一种基于旋翼无人机的植被背景下隐藏地物的探测方法,该方法通过搭载高光谱成像光谱仪的旋翼无人机,对目标区域进行逐行扫描,获得光谱数据并存储数据；同时采用荧光提取算法计算出目标区域的荧光值信息；再通过卫星获取目标地区的遥感高分辨率光学遥感图像,对步荧光值信息和遥感图像进行预处理得到荧光值信息；最后采用夫琅禾费暗线方法从荧光值信息中提取荧光探针,进而判定该目标区域是否存在隐藏地物。本发明的探测方法对植被伪装技术有很好的探测作用,在国防领域可以有着广泛的应用。而且探测的叶绿素荧光可用于指示植物细胞生理生化过程,及时准确反映植物光合作用的真实工作状态、诊断植物的健康状况等。(The invention discloses a detection method of hidden ground objects under a vegetation background based on a rotor unmanned aerial vehicle, which comprises the steps of scanning a target area line by line through the rotor unmanned aerial vehicle carrying a hyperspectral imaging spectrometer, obtaining spectral data and storing the data; meanwhile, fluorescence value information of the target area is calculated by adopting a fluorescence extraction algorithm; then, obtaining a remote sensing high-resolution optical remote sensing image of the target area through a satellite, and preprocessing the step fluorescence value information and the remote sensing image to obtain fluorescence value information; and finally, extracting the fluorescent probe from the fluorescence value information by adopting a Fraunhofer dark line method, and further judging whether the target area has hidden ground objects. The detection method of the invention has good detection effect on vegetation camouflage technology and can be widely applied in the field of national defense. And the detected chlorophyll fluorescence can be used for indicating the physiological and biochemical processes of plant cells, timely and accurately reflecting the real working state of plant photosynthesis, diagnosing the health condition of plants and the like.)

1. A detection method of hidden ground objects under a vegetation background based on a rotor unmanned aerial vehicle is characterized by comprising the following steps:

step S1: the rotor unmanned aerial vehicle carrying the hyperspectral imaging spectrometer is adopted, and the rotor unmanned aerial vehicle is hovered above a target area to be detected;

step S2: after the unmanned gyroplane is hovered, scanning a target area line by line through a hyperspectral imaging spectrometer, extracting data obtained by scanning by adopting a fluorescence extraction algorithm, obtaining fluorescence image data and storing the data;

step S3: acquiring a remote sensing high-resolution optical remote sensing image of the target area through a satellite, and preprocessing the fluorescence image data and the high-resolution optical remote sensing image obtained in the step S2 to obtain fluorescence value information, wherein the preprocessing comprises remote sensing image preprocessing, image registration, fusion, classification and non-vegetation image information elimination;

step S4: extracting a fluorescent probe from the preprocessed fluorescence value information by adopting a Fraunhofer dark line method, judging whether a hidden ground object exists in the target area according to the fluorescent probe, and if the fluorescence value of the fluorescent probe is less than 0.1mW m^-2nm^-1sr^-1If yes, judging that the target area has a hidden ground object; if the fluorescence value of the fluorescent probe is more than or equal to 0.1 and less than or equal to 2.5mW m^-2nm^-1sr^-1Then, the type of vegetation in the target area is manually identified through the high-resolution optical remote sensing image, the actual fluorescence parameter of the vegetation of the type is found out and compared with the fluorescence value of the fluorescence probe, and if the fluorescence probe accords with the corresponding parameter range, hidden ground objects do not exist; such as fluorescent probesIf the needle does not accord with the corresponding parameter range, a hidden ground object exists; if the fluorescence value of the fluorescent probe is more than 2.5mW m^-2nm^-1sr^-1If so, no hidden ground object exists.

2. The method of claim 1, wherein the method comprises the steps of: and step S2, extracting the data obtained by scanning by using a fluorescence 3FLD extraction algorithm, obtaining fluorescence image data, and storing the data.

3. The method of claim 1, wherein the method comprises the steps of: the hyperspectral imaging spectrometer is a Headwall high-resolution imaging spectrometer.

4. The method of claim 1, wherein the method comprises the steps of: the remote sensing image preprocessing comprises the steps of carrying out radiation correction, atmospheric correction and image cutting on data on the remote sensing high-resolution optical remote sensing image.

Technical Field

The invention belongs to the field of detection of hidden ground objects, and particularly relates to a detection method of hidden ground objects under a vegetation background based on a rotor unmanned aerial vehicle.

Background

The camouflage is that the original characteristic information of the military related target is changed by using the technical means of electronics, electromagnetism, optics, heat, acoustics and the like, the detection characteristic of the real target of the target is reduced or eliminated, and the 'invisibility' of the target is realized; or simulating the detection characteristics of the target, or simulating the detection characteristics of the target to imitate a false target to 'show false'. The basic methods of camouflage include natural camouflage, plant camouflage, camouflage color camouflage, artificial barrier camouflage, smoke camouflage and other technologies. Natural camouflage is the increase in the significance of hiding or reducing objects in the best of ground conditions, objects, night or poor visibility weather climates. The vegetation camouflage technology is a technology for camouflage of a target by using methods such as plant harvesting and the like. Camouflage is the use of dyes, paints or other materials to change the color of an object or to use camouflage patterns to reduce the difference between the object and the background and to change the external shape of the object. The artificial barrier camouflage is made of various camouflage materials, and is provided with a hidden barrier for a target, and mainly comprises a jungle camouflage color camouflage net, a multi-spectrum camouflage screen and a middle infrared band detection camouflage screen. The smoke camouflage mainly uses smoke to shield a target, and is used for detecting interference of an optical wave band of an opposite party.

At present, the self-adaptive camouflage technology (color-changing multispectral camouflage, camouflage materials with adjustable emissivity and temperature control, light broadband radar materials and the like) capable of naturally fusing the target and the natural background in all weather, all time periods and the whole process is mainly developed abroad. For example, aiming at the characteristics of remote sensing visible light and multispectral remote sensing detection, the self-adaptive camouflage material close to the spectrum of the target background ground object is developed on the basis of the original color camouflage by the camouflage material, and the target ground object is hidden by the material, so that the difficulty of detecting the target by the traditional remote sensing means is greatly increased. The jungle camouflage color and pure green camouflage net commonly used in vegetation environment has spectral characteristics similar to those of vegetation in visible light and near infrared partial areas. The traditional visible light remote sensing can intuitively reflect the information of the size, the shape, the size and the like of a ground object, and has great detection difficulty for a camouflage pattern target which changes the color and the brightness and is matched with the background color. Multispectral remote sensing can well identify typical ground objects such as vegetation, but the detection capability of camouflage materials close to the spectrum of a target background ground object is limited.

Chlorophyll fluorescence remote sensing is one of the leading research hotspots in the field of remote sensing earth observation. Chlorophyll fluorescence is a weak light emitted by plants and can indicate physiological and biochemical processes of plant cells. According to the characteristics of the satellite remote sensing vegetation fluorescent probe, the camouflage ground object with the vegetation as the background can be well detected and identified. A carbon satellite (TanSat) emitted in 2017 in China carries a hyperspectral imager special for detecting chlorophyll fluorescence, and can capture weak light rays emitted in the photosynthesis process of plants, namely chlorophyll fluorescence. Remote sensing via a survey satellite may collect a suitable data source for research.

At present, the chlorophyll fluorescence dynamics technology is widely applied to the fields of plant physiology, plant ecology and the like. Research means is mostly limited to leaf point contact measurement by means of adjusting a chlorophyll fluorescence instrument, although the mechanism of photosynthesis of plants can be revealed. Laser Induced Fluorescence (LIF) technology based on target fluorescence spectrum detection is generally used for detecting fluorescence spectrum signals emitted by plants in a certain distance range by using ultraviolet light of 320-380 nm as an excitation light source. However, researches show that the fluorescence spectrum characteristics (such as intensity, spectrum shape and relative intensity comparison of different wavelength regions) of plants are closely related to the excitation wavelength and performance of a detecting instrument, the types, concentrations, biomass and the like of pigments, and laser-induced fluorescence remote sensing has certain limitation in practical application. This limits to some extent the detection of concealed ground objects by means of fluorescence signals. Meanwhile, the laser induced fluorescence technology needs a large-capacity power supply device and is difficult to carry on an unmanned aerial vehicle under the condition.

Many scholars believe that to achieve spatial remote sensing of plant fluorescence it is feasible to rely on measuring the daylight-induced fluorescence signal. Although the signal of chlorophyll fluorescence is weak, the chlorophyll fluorescence information emitted by plants can be obtained from the reflectivity by methods such as vegetation index and solar fraunhofer dark line extraction. The fluorescence signal will be separated from the apparent reflection by an imaging spectrometer with high spectral resolution carried by the satellite. When plants are irradiated with light of a suitable wavelength, they fluoresce and, in the wavelength range of 660nm to 800nm, there are two distinct fluorescence peaks at 687nm and 760.5nm, respectively (Perez-Priego et al, 2005). Fluorescence in the red region is derived from chlorophyll within photosystem II, and near-infrared SIF is derived from chlorophyll within photosystem II and photosystem I (Meroni et al, 2009), as shown in figure 1. However, the vegetation index characteristics collected by satellite remote sensing alone cannot accurately detect and identify the hidden ground features under the vegetation background. Therefore, it is urgently needed to develop a method for detecting and identifying the camouflage material close to the target background ground object spectrum by using a remote sensing means.

Disclosure of Invention

The technical problem to be solved is as follows: the invention provides a detection method of hidden ground objects under a vegetation background based on a rotor unmanned aerial vehicle. The invention utilizes the satellite remote sensing vegetation probe to identify and detect the hidden target, provides a new detection mode for the hidden target taking vegetation as the background through a quantitative analysis method, and has great advantages in the aspect of detecting the target. Chlorophyll fluorescence is a weak light emitted by plants and can indicate physiological and biochemical processes of plant cells. According to the characteristics of the satellite remote sensing vegetation fluorescent probe, the camouflage ground object with the vegetation as the background can be well detected and identified.

The technical scheme is as follows: a detection method of hidden ground objects under a vegetation background based on a rotor unmanned aerial vehicle comprises the following steps:

step S1: the rotor unmanned aerial vehicle carrying the hyperspectral imaging spectrometer is adopted, and the rotor unmanned aerial vehicle is hovered above a target area to be detected;

step S2: after the unmanned gyroplane is hovered, scanning a target area line by line through a hyperspectral imaging spectrometer, extracting data obtained by scanning by adopting a fluorescence extraction algorithm, obtaining fluorescence image data and storing the data;

step S3: acquiring a remote sensing high-resolution optical remote sensing image of the target area through a satellite, and preprocessing the fluorescence image data and the high-resolution optical remote sensing image obtained in the step S2 to obtain fluorescence value information, wherein the preprocessing comprises remote sensing image preprocessing, image registration, fusion, classification and non-vegetation image information elimination;

step S4: extracting a fluorescent probe from the preprocessed fluorescence value information by adopting a Fraunhofer dark line method, judging whether a hidden ground object exists in the target area according to the fluorescent probe, and if the fluorescence value of the fluorescent probe is less than 0.1mW m^-2nm^-1sr^-1If yes, judging that the target area has a hidden ground object; if the fluorescence value of the fluorescent probe is more than or equal to 0.1 and less than or equal to 2.5mW m^-2nm^-1sr^-1Then, the type of vegetation in the target area is manually identified through the high-resolution optical remote sensing image, the actual fluorescence parameter of the vegetation of the type is found out and compared with the fluorescence value of the fluorescence probe, and if the fluorescence probe accords with the corresponding parameter range, hidden ground objects do not exist; if the fluorescent probe does not accord with the corresponding parameter range, a hidden ground object exists; if the fluorescence value of the fluorescent probe is more than 2.5mW m^-2nm^-1sr^-1If so, no hidden ground object exists.

Further, in step S2, the fluorescence 3FLD extraction algorithm is used to extract the data obtained by scanning, obtain fluorescence image data, and store the data.

Further, the hyperspectral imaging spectrometer is a Headwall high-resolution imaging spectrometer.

Further, the remote sensing image preprocessing comprises the steps of performing radiation correction, atmospheric correction and image cutting on data on the remote sensing high-resolution optical remote sensing image.

The mechanism of the invention is as follows: the chlorophyll fluorescence of the vegetation is very weak under natural light, and is only equivalent to 0.5% -2% of the light energy absorbed by the vegetation, however, due to the absorption of the solar spectrum by the atmosphere, the solar spectrum reaching the earth surface has a plurality of dark lines with the wave band width of 0.1-10, namely Fraunhofer absorption dark lines: hydrogen absorbs H formed at 656nm_αA dark line; oxygen molecules in the atmosphere absorb O formed near 760nm and 687nm₂-A dark line and O₂-B dark line. In the dark line wave band of fraunhofer absorption, the reflected light of the plants is very weak, the fluorescence is prominent, and the apparent reflectivity of the vegetation is larger than the real reflectivity.

Has the advantages that: the invention provides a rotary wing unmanned aerial vehicle-mounted hyperspectral imaging spectrometer based vegetation fluorescence detection method for detecting and identifying a camouflage target with vegetation as a background. The technical scheme has good detection effect on the vegetation camouflage technology and can be widely applied to the field of national defense. Meanwhile, chlorophyll fluorescence is weak light emitted by the plant, can indicate physiological and biochemical processes of plant cells, and can timely and accurately reflect the real working state of plant photosynthesis, diagnose the health condition of the plant and the like.

Drawings

Figure 1, a graph of daylight-induced chlorophyll fluorescence radiance and reflectance spectra (from Meroni et al, 2009) where (a) is the fluorescence radiance and (B) is the reflectance spectrum.

Fig. 2, a basic principle schematic diagram of a fluorescence algorithm (from Meroni et al, 2009), where (a) is the irradiance spectrum of the sun and (B) is the observed total radiance spectrum chlorophyll fluorescence and green vegetation reflectance spectra of the ground feature.

Fig. 3 is a schematic diagram of a hyperspectral imaging spectrometer on an unmanned aerial vehicle according to the present invention scanning a target area.

FIG. 4 is a schematic diagram of a process of calculating fluorescence values of each pixel in a measured spectral image line by using a fluorescence 3FLD extraction algorithm to obtain fluorescence value characteristics related to the whole image.

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings.