A method for retrieving the concentration of gas pollutants under complex backgrounds based on infrared hyperspectral

Abstract

The present invention discloses a method for retrieving the concentration of gas pollutants under a complex background based on infrared hyperspectral, which relates to the field of infrared hyperspectral remote sensing. The problem to be solved is the inversion of temperature emissivity of thermal infrared hyperspectral data. The steps of the method are as follows: (1) Acquisition of infrared hyperspectral data and its auxiliary data; (2) Calculation of atmospheric radiation transfer parameters; (3) Atmospheric influence correction; (4) Determination of gas pollutant types; (5) Inversion of background temperature and emissivity. The present invention gives full play to the advantages of infrared hyperspectral technology, adopts a strategy of inversely retrieving different parameters in stages, and improves the inversion accuracy of gas pollutant concentration under a complex background by improving the solving accuracy of background temperature and emissivity.

Description

Technical Field

[0001] The present invention relates to the field of infrared hyperspectral remote sensing, and particularly to a method for retrieving the concentration of gaseous pollutants under complex infrared hyperspectral backgrounds. Background Art

[0002] Gaseous pollutants are diffused and flowing in the air, without a fixed shape or volume, and especially rapidly diffuse outward with the circulation of air. Moreover, most gases in the air are colorless or difficult to identify with the naked eye. Many chemically flammable, explosive, and harmful gases can cause combustion or damage to human organs even at very low concentrations, which brings great difficulties to the discovery and real-time monitoring of gaseous pollutants. For the detection of gaseous pollutants, many attempts have been made by experts and scholars at home and abroad, and real-time monitoring using instruments has become almost the only monitoring method. Electrochemical, semiconductor, gas chromatography, thermal, magnetic, and optical instruments are all commonly used gaseous pollutant detection devices. Infrared imaging technology is a type of optical gaseous pollutant detection device and is very suitable for large-scale and long-distance detection. Compared with traditional thermal imagers, infrared hyperspectral imaging has the characteristics of combining spectra and images, with multiple bands and rich information content, and is considered to have unique advantages in gaseous pollutant detection. More than 95% of gaseous pollutants have such "fingerprint" characteristics, which is a favorable method for identifying gaseous pollutants.

[0003] Although the inversion of gaseous pollutants and concentration has the support of clear remote sensing physical models and spectral analysis models, due to the lack of measurement data, it is difficult to determine parameters such as background radiation brightness spectra, atmospheric parameters, and target gas temperatures. Therefore, the present invention proposes a method for retrieving the concentration of gaseous pollutants under complex infrared hyperspectral backgrounds. On the one hand, this method gives full play to the advantages of high spectral resolution and a large number of measurements of infrared hyperspectra, moderately uses prior information, and inversely calculates as many parameters as possible based on measurement data; on the other hand, it uses reliable measurement data and atmospheric radiation transfer software to achieve more accurate estimation of atmospheric parameters. Summary of the Invention

[0004] Aiming at the existing technical gaps and deficiencies, the technical problem to be solved by the present invention is to provide a method for retrieving the concentration of gaseous pollutants with higher accuracy under complex backgrounds.

[0005] To solve the above technical problem, the present invention provides a method for retrieving the concentration of gaseous pollutants under complex infrared hyperspectral backgrounds. The specific technical solutions adopted are as follows:

[0006] (1) Acquisition of infrared hyperspectral data and its auxiliary data. Use an infrared hyperspectral imager to image the suspected gaseous pollution area to obtain infrared hyperspectral data in the range of 3 - 5 μm and 7.0 - 13.0 μm. Simultaneously measure the air temperature and humidity;

[0007] (2) Calculation of atmospheric radiation transfer parameters. Input the collected relative humidity, air temperature, geographical location, and time into the atmospheric radiation transfer model, and calculate the corresponding atmospheric background radiation spectrum (L a ), path transmittance spectrum (t a ), and path radiance spectrum (L p ) through the atmospheric radiation transfer model;

[0008] (3) Atmospheric influence correction. Remove the atmospheric influence according to the corresponding atmospheric background radiation spectrum, path transmittance spectrum, and path radiance spectrum calculated by the atmospheric radiation transfer model. The calculation method is

[0009]

[0010] where L gas is the radiance leaving the background through the gas pollutant; L is the radiance entering the pupil received by the infrared hyperspectral; λ k is the central wavelength of the k-th channel of the infrared spectral data, k = 1, 2, 3,..., N k ; N k is the number of channels;

[0011] (4) Determination of gas pollutant types. According to prior knowledge, the temperature of the gas pollutant is divided into two cases: high temperature and non-high temperature. If the temperature of the gas pollutant is 10°C or more higher than the gas temperature, the temperature of the gas pollutant is considered high temperature. Otherwise, the temperature of the gas pollutant is considered normal temperature. Perform range normalization transformation on the radiance spectrum leaving the gas to obtain the normalized radiance spectrum leaving the gas. Perform Logistic function transformation on the normalized radiance spectrum leaving the gas. The transformation method is

[0012]

[0013] where x is the input value; y is the input value. Remove the envelope line from the radiance spectrum after the Logistic function transformation, and extract the absorption wavelengths to form a set of absorption wavelengths. Compare the set of absorption wavelengths with the gas pollutant spectral library to determine the gas pollutant type;

[0014] (5) Inversion of background temperature and emissivity. Select the non-gas pollutant absorption band in the infrared hyperspectral data, and combine it with the atmospheric background radiation spectrum (L a ), and use the temperature-emissivity separation algorithm to calculate the background temperature (T) and emissivity (ε) for each pixel;

[0015] (6) Inversion of gas pollutant concentration. Search for the spectrum most similar to the background pixel in the emissivity spectrum library of the ground object, interpolate the emissivity curve of the background pixel to obtain the emissivity value in the absorption band of the gas pollutant, and obtain the complete emissivity spectrum of the background pixel. Extract the absorption coefficient spectrum (α gas ) of the gas pollutant from the gas pollutant spectrum library. Take the gas pollutant concentration (c gas ) and the gas temperature (T gas ) as the target variables, and establish a cost function

[0016]

[0017]

[0018] where σ is the cost function; B(·) is the Planck function; θ is the observation normal angle of the pixel. Use the optimization algorithm to find the optimal values of the gas pollutant concentration and the gas temperature for each pixel. Take the optimal value of the gas pollutant concentration as the gas pollutant concentration of the pixel.

[0019] The beneficial effects of the present invention are as follows: The present invention makes full use of the temperature emissivity separation and spectral transformation algorithms, adopts the strategy of inversely retrieving different parameters in stages, is more effective for the inversion of gas pollutant concentration under complex backgrounds, and can further improve the inversion accuracy. Description of the Drawings

[0020] Figure 1 Flow chart of the method for inverting gas pollutant concentration under complex backgrounds based on infrared hyperspectral Detailed Embodiments

[0021] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not limited to the present invention. Any similar methods and their similar variations using the present invention should be included in the protection scope of the present invention.

[0022] As Figure 1 shown, a method for inverting temperature emissivity of thermal infrared hyperspectral based on spectral resolution degradation includes the following steps:

[0023] (1) Acquisition of infrared hyperspectral data and its auxiliary data. Use an infrared hyperspectral imager to image the suspected gas pollution area to obtain infrared hyperspectral data in the range of 3-5 μm and 7.0-13.0 μm. Synchronously measure the air temperature and humidity;

[0024] (2) Calculation of atmospheric radiation transfer parameters. Input the collected relative humidity, air temperature, geographical location, and time into the atmospheric radiation transfer model, and calculate the corresponding atmospheric background radiation spectrum (L a )、path transmittance spectrum (t a ) and path radiance spectrum (L p ) through MODTRAN;

[0025] (3) Atmospheric influence correction. Remove the atmospheric influence according to the calculated atmospheric background radiation spectrum, path transmittance spectrum, and path radiance spectrum from the atmospheric radiation transfer model. The calculation method is

[0026]

[0027] where L gas is the radiance of the gas leaving the background through the gas pollutant; L is the radiance of the incoming pupil received by the infrared hyperspectral; λ k is the central wavelength of the k-th channel of the infrared spectral data, k = 1, 2, 3,..., N k ; N k is the number of channels;

[0028] (4) Determination of gas pollutant types. According to prior knowledge, the temperature of gas pollutants is divided into two cases: high temperature and non-high temperature. If the temperature of the gas pollutant is 10°C or more higher than the gas temperature, the temperature of the gas pollutant is considered high temperature. Otherwise, the temperature of the gas pollutant is considered normal temperature. Perform range normalization transformation on the radiance spectrum of the gas leaving the background to obtain the normalized radiance spectrum of the gas leaving the background. Perform Logistic function transformation on the normalized radiance spectrum of the gas leaving the background. The transformation method is

[0029]

[0030] where x is the input value; y is the input value. Remove the envelope from the radiance spectrum after the Logistic function transformation and extract the absorption wavelengths to form a set of absorption wavelengths. Compare the set of absorption wavelengths with the gas pollutant spectral library to determine the gas pollutant type;

[0031] (5) Inversion of background temperature and emissivity. Select the non-gas pollutant absorption band in the infrared hyperspectral data, and combine it with the atmospheric background radiation spectrum (L a ), and use the temperature emissivity separation algorithm to calculate the background temperature (T) and emissivity (ε) for each pixel;

[0032] (6) Inversion of gas pollutant concentration. Search for the spectrum most similar to the background pixel in the emissivity spectrum library of the ground object, and interpolate the emissivity curve of the background pixel to obtain the emissivity value in the absorption band of the gas pollutant, thus obtaining the complete emissivity spectrum of the background pixel. Extract the absorption coefficient spectrum (α gas ) of the gas pollutant from the gas pollutant spectrum library. Take the gas pollutant concentration (c gas ) and the gas temperature (T gas ) as the target variables to establish a cost function

[0033]

[0034]

[0035] where σ is the cost function; B(·) is the Planck function; θ is the observation normal angle of the pixel. Use the optimization algorithm to find the optimal values of the gas pollutant concentration and the gas temperature pixel by pixel. Take the optimal value of the gas pollutant concentration as the gas pollutant concentration of this pixel.

Claims

1. A method for retrieving the concentration of gas pollutants under complex infrared hyperspectral backgrounds, characterized in that it includes the following steps: (1) Acquisition of infrared hyperspectral data and its auxiliary data. An infrared hyperspectral imager is used to image the suspected gas pollution area to obtain infrared hyperspectral data in the range of 3 - 5 μm and 7.0 - 13.0 μm, and the air temperature and humidity are measured synchronously; (2) Calculation of atmospheric radiation transfer parameters. Input the collected relative humidity, air temperature, geographical location, and time into the atmospheric radiation transfer model, and calculate the corresponding atmospheric background radiation spectrum L a , path transmittance spectrum t a and scattered radiation spectrum L p ; (3) Atmospheric influence correction. According to the atmospheric radiation transfer model, the corresponding atmospheric background radiation spectrum, path transmittance spectrum, and scattered radiation spectrum are calculated to remove the atmospheric influence. The calculation method is: Among them, Lgas is the outgoing gas radiance leaving the background through gas pollutants; L is the incoming pupil radiance received by the infrared hyperspectral, and λ k is the central wavelength of the k-th channel of the infrared spectral data, where k = 1, 2, 3,..., N k ; N k is the number of channels; (4) Determination of gas pollutant types. Based on prior knowledge, the temperature of gas pollutants is divided into two cases: high temperature and non - high temperature. If the temperature of the gas pollutant is 10 °C or more higher than the gas temperature, the temperature of the gas pollutant is considered high temperature; otherwise, the temperature of the gas pollutant is considered normal temperature. The radiance spectrum of the off - gas is subjected to range normalization transformation to obtain the normalized radiance spectrum of the off - gas, and the Logistic function transformation is performed on the normalized radiance spectrum of the off - gas. The transformation method is: where x is the input value; y is the input value; After the Logistic function transformation, the envelope of the radiance spectrum is removed, and the absorption wavelengths are extracted to form a set of absorption wavelengths. The set of absorption wavelengths is compared with the gas pollutant spectral library to determine the type of gas pollutant; (5) Background temperature and emissivity inversion. Select the absorption bands of non-gas pollutants in the infrared hyperspectral data, and combine with the atmospheric background radiation spectrum L a , and use the temperature-emissivity separation algorithm to calculate the background temperature T and emissivity ε for each pixel; (6) Gas pollutant concentration inversion. Search for the spectrum most similar to the background pixel in the emissivity spectrum library of ground objects, and interpolate the emissivity curve of the background pixel to obtain the emissivity value in the absorption band of gas pollutants, thus obtaining the complete emissivity spectrum of the background pixel; extract the absorption coefficient spectrum α of gas pollutants from the gas pollutant spectrum library. gas ; Take the gas pollutant concentration c gas and the gas temperature T gas as target variables to establish a cost function: where σ is the cost function; B(·) is the Planck function; θ is the observation normal angle of the pixel; The optimal values of the gas pollutant concentration and gas temperature are obtained pixel - by - pixel using the optimization algorithm, and the optimal value of the gas pollutant concentration is used as the gas pollutant concentration of the pixel.

Images