CN111272664B - Synchronous correction method for field measurement spectrum of geophysical spectrometer - Google Patents
Synchronous correction method for field measurement spectrum of geophysical spectrometer Download PDFInfo
- Publication number
- CN111272664B CN111272664B CN202010109585.6A CN202010109585A CN111272664B CN 111272664 B CN111272664 B CN 111272664B CN 202010109585 A CN202010109585 A CN 202010109585A CN 111272664 B CN111272664 B CN 111272664B
- Authority
- CN
- China
- Prior art keywords
- spectrum
- formula
- data
- spectrometer
- fitting
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
Abstract
The invention discloses a synchronous correction method for a field measurement spectrum of a geological spectrometer. The main influencing factor of the spectral data is derived from the asynchrony between the reflected and incident radiation intensities, so that the weather stability requirements during the measurement are strict. The invention is as follows: 1. and establishing a fitting model of the incident radiance and the illumination intensity. 2. And synchronously correcting the reflection spectrum of the ground object. The invention provides a new ground object spectrum measurement method by establishing the model of the incident radiance and the illumination intensity, avoids repeatedly calibrating the incident radiance after the environment changes, and simplifies the test. The invention is applied to the field work of measuring the spectrum of the ground feature by the ground feature spectrometer, can solve the problem that the collected spectrum generates certain spectrum error due to illumination change caused by cloudy weather or other factors, and increases the applicable weather range of the ground feature spectrometer.
Description
Technical Field
The invention belongs to the technical field of field spectrum measurement, and particularly relates to a synchronous correction method for field measurement of a ground object spectrum by a ground object spectrometer.
Background
In order to study the reflection spectrum of visible and near-infrared bands of various ground objects under the natural conditions in the field and grasp the spectral radiation characteristics of various ground objects on the ground, a large amount of ground object spectrum observation and study are needed. Remote sensing reflectivity R of ground object when the illumination wavelength is lambda rs (λ)＝L w (λ)/E d (lambda), i.e. the reflection radiance L of the measured object under the illumination of lambda wavelength w Incident irradiance E on the surface of the object d The ratio of the first to the second.
The feature spectrometer is a common instrument for measuring the spectrum of the feature, and a standard white board (1/Π in reflectivity) is required to pass through in the process of measuring the spectrum of the feature by using the spectrometer
white panel) for white panel calibration, and the calibration result is used as the incident radiance E for measuring the reflectivity d Then aiming at the ground object to be measured to obtain the reflected radiance L w And obtaining the reflectivity according to the ratio of the two. Therefore, the main influence factor of the spectrum data comes from the asynchronism between the reflected radiance and the incident radiance, so that the requirement on the weather stability is strict in the measuring process, and great inconvenience is brought to the continuous measuring process of the ground object spectrum in cloudy weather and other climates with quick illumination change. In southern areas of china, the acquisition of remote sensing data is also very inconvenient for climatic reasons during the year. According to the meteorological records, during the period from 2010.1.1 to 2019.10.1, the sunny weather in the Hangzhou region occupies 12% of the total days, the cloudy weather occupies 40% of the total days, so the date for collecting the spectrum has great limitation, and is not beneficial to the continuous spectrum collection work.
The spectral range measured by the existing terrestrial object spectrometer reaches 350-2500nm, contains visible light and infrared bands, and has better spectral resolution (1 nm), but field measurement can be influenced by water vapor absorption. The water vapor absorption is the absorption effect of water vapor in the atmosphere on electromagnetic radiation with a certain wavelength, most of the water vapor is concentrated on an infrared band, so that the electromagnetic radiation with the corresponding wavelength cannot be received on the ground, and certain difficulty is brought to quantitative analysis of incident radiation brightness measured by a ground object spectrometer, so that the method carries out synchronous correction on the spectrum measured by the ground object spectrometer in a visible light-near infrared (VNIR) band.
The problem of synchronization between incident radiance and reflection radiance in the spectral measurement process is solved, the quality of obtaining the spectrum can be effectively improved, the operation convenience is improved, spectrum collection work can be carried out under various climates, and the synchronization problem in the spectral measurement process is not solved through research at present.
Disclosure of Invention
The invention aims to provide a method for simulating the incident radiance in a visible light-near infrared band range (VNIR: 400-900 nm) by illumination intensity, calculating the spectral reflectivity according to the simulated incident radiance and quantifying the error between the corrected spectrum and the original standard measurement spectrum, aiming at the difference of the incident radiance and the reflected radiance in the time of the spectral reflectivity composition obtained in the measurement process of the existing near-earth spectrometer and the limitation that the spectrum acquisition can only be carried out in the stable illumination weather at present.
The method comprises the following specific steps:
step one, establishing a fitting model of incident radiance and illumination intensity
1-1. Building data sets
And detecting the incident radiance spectrum once or for many times through a standard white board and a ground object spectrometer under different illumination intensities to obtain a comprehensive data set. The comprehensive data set is split into a training set and a testing set.
1-2, establishing a fitting model
Fitting the data of each wave band of the incident radiance spectrum in the training set respectively to obtain the incident radiance Rad of each wave band m And the illumination intensity lux, as shown in formula (1), and the wavelength m =400,401, \8230;, 900.
Rad m ＝f m (lux) formula (1)
In the formula (1), f m As a function of the intensity of the incident radiation at m wavelengths and the intensity of the illumination.
1-3, verifying the fitting function through the data of the test set
Respectively substituting the illumination intensity data in the test set data into fitting functions shown in formula (1), and calculating illumination intensity lux 'of the ith sample in the test set' i Corresponding predicted incident radiance spectral data Rad i (prediction) as shown in formula (2), i =1,2, \ 8230;, n. And n is the number of samples in the test set.
Rad i (prediction) = (f) 400 (lux′ i )，f 401 (lux′ i )，…，f 900 (lux′ i ) C type (2)
Calculating fitting accuracy ref of fitting model im The expression (b) is shown in formula (3). ref (r) f im And (4) representing the fitting accuracy of the fitting model to the ith sample of the test set under the m wave band. i =1,2, \8230;, n; m =400,401, \ 8230;, 900.
In formula (3), rad im (true) is the true incident radiance at m-band for the ith sample of the test set; rad im (prediction) incident radiance is predicted at m-band for the ith sample of the test set.
Calculating the root mean square error RMSE of the fitting function m Is represented by the formula (4).
If RMSE m And if the value is higher than the preset value, the step 1-1 is executed again, and the fitting function is updated.
Step two, synchronous correction of ground object reflection spectrum
And collecting the illumination intensity lux' in the environment while collecting the brightness of the reflection radiation of the ground object by the ground object spectrometer. The calculated and predicted incident radiance spectrum Rad (in) is shown as equation (5).
Rad (in) = (f) 400 (lux″)，f 401 (lux″)，…，f 900 (lux″)) Into ＝(Rad 400 into ,Rad 401 into …,Rad 900 go into ) Formula (5)
Calculating a synchronous corrected reflectance spectrum ref of a surface feature Synchronization As shown in equation (6).
In the formula (6), rad (reverse) is a ground object reflection radiance spectrum of 400-900nm wave band obtained by measurement of a ground object spectrometer.
Preferably, in step 1-1, the resulting spectrum of incident radiance comprises the value of incident radiance per 1nm wavelength in the 400-900nm band.
Preferably, in the step 1-1, the training set accounts for 2/3 of the total amount of data in the comprehensive data set; the test set accounts for 1/3 of the total amount of data in the integrated data set.
Preferably, the Field SPEC4 portable Field spectrograph is adopted as the Field spectrograph; the illuminometer adopts an ST-85 illuminometer.
The invention has the beneficial effects that:
1. compared with the existing field ground object spectrum measurement method for obtaining the ground object spectrum by respectively acquiring the white board reflected radiance and the ground object reflected radiance and calculating, the invention provides a new ground object spectrum measurement method by establishing the model of the incident radiance and the illumination intensity, avoids repeatedly calibrating the incident radiance after the environment changes, and simplifies the measurement process.
2. The invention is applied to the field work of measuring the spectrum of the ground feature by the ground feature spectrometer, can solve the problem that the collected spectrum generates certain spectrum error due to illumination change caused by cloudy weather or other factors, and increases the applicable weather range of the ground feature spectrometer.
Drawings
FIG. 1 is a graph showing the variation of the correlation coefficient of the fitting function in each band;
FIG. 2 is a test set validation result of the present invention;
FIG. 3 is a graph of RMSE curves for each band of the incident radiance spectrum of the present invention;
FIG. 4 is a comparison of the synchronous calibration spectrum and the standard measurement spectrum of the soil;
FIG. 5 is a comparison of the synchronous correction spectrum of the vegetation with the standard measurement spectrum;
FIG. 6 is a comparison of the synchronous calibration spectrum and the standard measurement spectrum of a water body;
FIG. 7 is a diagram of relative error analysis of the simultaneous calibration spectrum and the standard spectrum of three features.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
A synchronous correction method for field measurement spectrum of a geological spectrometer comprises the following specific steps:
step one, establishing a fitting model of incident radiance and illumination intensity: and modeling and evaluating the precision according to all incident radiance data and corresponding illumination intensity data acquired by weather conditions such as sunny days, cloudy days and the like, wherein the illumination intensity range comprises an illumination condition range suitable for daily spectral measurement.
1-1. Building data sets
And detecting the incident radiance spectrum once or for many times through a standard white board and a ground object spectrometer under different illumination intensities to obtain a comprehensive data set. The incident radiance spectrum contains the value of the incident radiance per 1nm wavelength in the 350-2500nm band. Each sample in the integrated dataset contains one illumination intensity data and incident radiance spectral data corresponding thereto. In order to evaluate the accuracy of the model, 2/3 of data in the comprehensive data set is randomly selected as a training set to establish the model, and the rest 1/3 of data are used as a testing set to verify the model. Since the incident radiances of different wavelengths are different, it is necessary to separately model the illumination intensity data in the training set and the incident radiance data of each corresponding wavelength. Meanwhile, because the water vapor absorption and the edge distortion of the instrument can generate larger influence on the data acquired by the instrument, the invention only corrects the visible light-near infrared band (400-900 nm).
The feature spectrometer is a Field Spec4 portable feature spectrometer (ASD spectrometer for short) produced by ASD (Analytical Spectral Devices) of America, and has a Spectral measurement range of 350-2500nm, a wave band resolution of 1nm and a Field angle of 15 degrees. The illuminometer for measuring the illumination intensity is an ST-85 illuminometer produced by a photoelectric instrument factory of Beijing university, the illumination intensity range is 0-200000lux, the sampling interval is 0.3s, and the light intensity range can be automatically selected according to the illumination intensity change. ST-85 adopts a plane type photosensitive sensor, is different from a spherical photosensitive sensor which is used more, can effectively improve the shadow change caused by the change of the sun angle, and reduces the influence of the sun angle on the illumination data.
1-2, establishing a fitting model
And (3) performing correlation analysis on the illumination intensity data and the incident radiance data of each wavelength, so as to analyze the correlation relationship of the illumination intensity data and the incident radiance data of each wavelength, wherein the linear relationship of the illumination intensity data and the incident radiance data of each wavelength is proved to be good as the correlation coefficient approaches to 1. According to the correlation analysis result, regression analysis is carried out, wherein the illumination intensity is used as independent variable, the incident radiance of each wave band is respectively used as dependent variable, and a fitting model of the illumination intensity and the incident radiance of each wave band is established through a least square method, so that the incident radiance Rad of each wave band can be obtained m And the illumination intensity lux, as shown in formula (1), and the wavelength m =400,401, \8230;, 900.
Rad m ＝f m (lux) formula (1)
In the formula (1), lux is the intensity of illumination, rad m Is the value of the m-wavelength incident radiation intensity, f m As a function of the intensity of the incident radiation at m wavelengths and the intensity of the illumination.
And 1-3, verifying the fitting function through the test set data, and evaluating the accuracy of the fitting model according to a verification result.
Respectively substituting the illumination intensity data in the test set data into the fitting function of the incident radiance and the illumination intensity of each wave band shown in the formula (1), recording the number of the samples in the test set as n, and then testing the illumination intensity lux in the ith sample in the test set i ' corresponding predicted incident radiance spectral data Rad i (prediction) as shown in formula (2), i =1,2, \ 8230;, n.
Rad i (prediction) = (f) 400 (lux′ i )，f 401 (lux′ i )，…，f 900 (lux′ i ) Formula (2)
In the formula (2), i is a sample number; f. of m (m = 400-900) is a fitted function of the intensity of the incident radiation and the intensity of the illumination, rad, for each wavelength i (prediction) is a predicted incident radiance spectrum obtained by fitting a function to the intensity of illumination in sample i.
In the remote sensing of ground features, the radiance is due to the external lightThe light environment changes, and the reflectivity spectrum of the ground object is generally used as an index for reflecting the remote sensing property of the ground object, so that the radiance data is converted into the reflectivity data, and a model can be better evaluated from the remote sensing angle of the ground object. According to the real standard white board reflection radiance spectrum Rad measured in the test set i (trues) and the calculated predicted incident radiance spectrum Rad i (prediction) comparison is carried out to obtain fitting accuracy ref of a fitting model im The expression (b) is shown in formula (3). ref (r) ref im And (4) representing the fitting accuracy of the fitting model to the ith sample of the test set under the m wave band. i =1,2, \8230;, n; m =400,401, \ 8230;, 900.
In formula (3), rad im (true) is the true value of the incident radiation brightness of the ith sample of the test set under the m wave band; rad im (prediction) as a function f m Predicting the incident radiance of the ith sample of the test set under the m wave band; .
The accuracy of the fitting model is verified through the difference between the predicted reflectivity spectrum and the standard reflectivity spectrum of the standard whiteboard, and meanwhile, as the fitting model has a fitting function at each wavelength, the function of each wavelength needs to be evaluated. The standard white board has a standard reflectance of 1 at each wavelength, n samples at each wavelength, and a Root Mean Square Error (RMSE) as an evaluation index. Root mean square error
Wherein n is the total number of samples, y i Is the true value of the sample i,
is the predicted value of sample i.
With the invention, the total number of samples is n; the true value is the standard accuracy spectrum of the standard white board, each wavelength is 1, the predicted value is the predicted reflectivity spectrum of the standard white board, and each wavelength passes through the corresponding simulated reflectivity spectrumAnd (4) performing resultant function calculation. The root mean square error RMSE of the fitting function for the wavelength m m Is represented by the formula (4).
In the formula (4), the reaction mixture is,
is constant at 1.
And calculating the RMSE of each wavelength test set in the 400-900nm wave band, namely evaluating the fitting precision of each wavelength in the fitting model. Root mean square error RMSE of fitting function m And if the value is higher than the preset value, the step 1-1 is executed again, and the fitting function is updated.
Step two, synchronous correction of ground object reflection spectrum
And according to the definition of the reflection spectrum of the ground object, obtaining the synchronous correction spectrum of the ground object based on the fitting model of the incident radiance and the illumination intensity. According to the fitting model established in the first step, the incident radiance value of each wavelength in the corresponding 400-900nm wave band under any illumination intensity condition can be obtained, so that the synchronous correction reflection spectrum of the ground object can be calculated by measuring the reflection radiance and the synchronous illumination intensity of the ground object. Meanwhile, the accuracy of the synchronous correction method can be verified by comparing the standard measurement spectrum of the ground object.
2-1. Synchronous correction of ground object reflection spectrum
And simultaneously acquiring the brightness of the reflection radiation of the ground object by the ground object spectrometer, and acquiring the illumination intensity lux' in the environment. And obtaining a corresponding predicted incident radiance spectrum Rad (input) based on the fitting model of the incident radiance and the illumination intensity as shown in a formula (5).
Rad (in) = (f) 400 (lux″)，f 401 (lux″)，…，f 900 (lux″)) Go into ＝(Rad 400 into ,Rad 401 into …,Rad 900 go into ) Formula (5)
In formula (5), f m (m = 400-900) is the approximation of the intensity of the incident radiation and the intensity of the illumination for each wavelengthA resultant function.
Defining R according to the remote sensing reflectivity of the surface features rs ＝L w /E d Calculating a synchronous corrected reflectance spectrum ref of the surface feature Synchronization As shown in equation (6).
In the formula (6), rad (reverse) is a 400-900 frequency band ground object reflection radiance spectrum measured by a ground object spectrometer.
2-2. Verification of synchronous correction spectrum of ground object
The verification of the synchronous correction spectrum of the ground object is evaluated by the relative error of the reflectivity of each wave band between the synchronous correction spectrum and the standard spectrum, so as to illustrate the relative deviation of the synchronous correction spectrum of the ground object at each wavelength. Wherein the standard spectrum of the ground feature is determined by standard spectral measurement of the ground feature. Relative error
Where δ is the relative error, a is an approximation, and a is the true value.
Therefore, the synchronous correction relative error of the m-wavelength ground object spectrum can be obtained:
wherein, the first and the second end of the pipe are connected with each other,
and
the reflectance and the standard reflectance are respectively corrected for the synchronization of the m-wavelength feature. The standard reflectivity is obtained by simultaneously collecting the incident radiance and the reflected radiance. The applicability of the synchronous correction method of the field surface feature spectrum is verified through calculation and analysis of relative errors of each wavelength of 400-900 nm.
To verify the accuracy of the invention, the following tests were performed:
1. data acquisition
Experiment (1) data acquisition of synchronous correction model
The experiment aims to obtain incident radiance and corresponding illumination intensity data within the illumination intensity range of 10000-100000lux, the incident radiance is obtained by measuring the reflection radiance of a standard white board through an ASD spectrometer, the illumination intensity is obtained by measuring through an ST-85 illuminometer, and the specific experiment steps are as follows:
1. the optical fiber of the spectrometer is placed in an optical fiber handle, the optical fiber handle is fixed on a tripod through a bottom screw, the tripod is leveled to ensure that the optical fiber is vertically downward, and a base is horizontally placed under the optical fiber.
2. And (3) placing the standard white board on the base and moving the standard white board to a position right below the optical fiber, and ensuring no shadow on the white board in the measurement process. The probe of the ST-85 illuminometer is placed beside the whiteboard, remaining in the same plane as the whiteboard and facing skyward for measuring the illuminance value.
3. Before data acquisition, the instrument needs to be optimized and parameter set, the spectrometer is started to be preheated for 15 minutes and then is connected with a computer, the type of the measured data is set to be 'raw DN', and the reflection radiance of the ground object is measured. The spectrum is optimized to collect the dark current in the period of stable weather, and the dark current needs to be collected again after the instrument is used for about 15 minutes. The sampling interval and the sampling times are set (each group of experiments does not exceed 15 min).
4. And (4) turning on the illuminometer, connecting the illuminometer with a computer, setting the sampling frequency to be 1 time/1 s, and recording continuous illumination data.
5. Firstly, the illumination data is measured, then the reflection radiance data of the white board is measured, after each group of experimental measurement of the ASD spectrometer is finished, the work of the illumination meter is stopped, and the data are respectively stored.
Experiment (2) synchronous correction model verification data acquisition of ground object
The purpose of the experiment is to obtain standard spectrum and reflection radiance data of the ground features under different conditions, so that the instrument used for verifying the applicability of the synchronous correction model in the spectrum measurement of different ground features is consistent with that in the first experiment. According to the common ground objects in the hyperspectral research, four common ground objects of mineral substances, soil, vegetation and water are selected for the experiment.
Selecting standard white board consistent with the white board in the step one, and selecting canopy leaves with plant names, tap water and ground CaSO from four ground objects respectively 4 The powder and the ground and dried soil powder are used as verification materials of vegetation, water, mineral substances and soil. In order to avoid the influence of different leaves in the plant growth process, the same leaf of the same plant of vegetation is selected as a measuring vegetation ground object in multiple verification experiments, and the leaf is flattened on black background cloth through a standard transparent glass sheet in the measuring process so as to avoid the difference in spectrum caused by the irregularity of the plant leaf. Tap water and CaSO 4 Powder and soil powder all hold in the pp plastic case, and wherein the inboard box wall of running water box all scribbles black with black water pen, avoids external light to disturb, and accurate measurement water spectrum, soil powder shine moisture behind abundant illumination to avoid the different influence to the soil spectrum of water content. The specific steps of the experiment are as follows:
1. and (4) installing experimental instruments, wherein the steps and the standards are consistent with those in the first experiment. After the ASD spectrometer is preheated for 15min, firstly, the illuminometer is opened to record the illumination intensity in the whole experiment process, and then different ground object data acquisition of the ground object verification experiment is carried out.
2. In the time period of stable illumination, after the spectrometer is optimized and the white board is calibrated, the type of the measured data is set as 'reflection', the reflectivity of the white board is measured, then the measured data is changed into 'raw DN', and the reflected radiance data of the white board is measured.
3. Placing a ground object on the base, moving the base to a position right below the optical fiber of the spectrograph, setting the type of the measured data as 'reflection', measuring the reflectivity of the ground object, then changing the measured data into 'raw DN', and measuring the reflection radiance data of the ground object.
4. And step 2 and step 3 are used for acquiring the once verification data of the ground features, and the verification data of each ground feature is acquired for multiple times under different illumination conditions.
2. Synchronous correction modeling:
as shown in FIG. 1, correlation analysis is performed on incident radiance and illumination intensity of each wave band, good correlation is shown in each wave band from 400nm to 900nm, the lowest correlation obtained at the 400nm wave band is 0.978, the correlation at the 576nm wave band is the highest and reaches 0.999, and the linear correlation characteristic between the incident radiance and the illumination intensity is very good. According to the linear characteristics of the incident radiance and the illumination intensity, a fitting model of the incident radiance and the illumination intensity under each wave band is established through unary linear regression, the precision analysis is carried out on each wave band model, and the result shows that the models of 400nm and 576nm wave bands, R 2 Respectively 0.9588 and 0.989, and the establishment of the linear regression model is proved to be capable of well fitting, and the model of each wave band can meet the precision requirement.
And (3) synchronously correcting the data of the test set according to the established model to obtain a series of corrected spectrums under different illumination conditions, and as can be seen from figure 2, the relative error between the reflectivity of each spectrum in a VNIR wave band and the standard reflectivity of 1 is within +/-0.05. And calculating the RMSE between the synchronous correction spectrum and the standard reflection spectrum of the test set, wherein the RMSE at a 400nm waveband is the highest and reaches 0.035, and the RMSE at a 626nm waveband is the lowest and is only 0.0097. The verification result of the test set is shown in fig. 3, which shows that the relative error of the synchronous correction spectrum meets the requirement, and the maximum RMSE is only 0.035, which proves that the maximum error of the model to the spectrum correction is also within the range of ± 5%, and the accuracy requirement can be met.
3. And (3) ground object verification of the model:
and (3) obtaining a synchronous correction spectrum of the ground object according to the ground object reflection radiance data and the corresponding illuminance value collected in the experiment (2) and the established synchronous correction model. The synchronous correction spectrums of different surface features and the corresponding surface feature standard spectrums are analyzed, and error quantification can be further carried out on the surface feature verification effect of the synchronous correction model.
As shown in fig. 4, 5 and 6, according to the synchronous correction spectrum and the standard measurement spectrum of the three ground features, it can be seen that the curves of the two are basically coincident, and a better consistency is presented. The error analysis of the synchronous correction result is shown in fig. 7, and it can be seen that the relative error of each wave band of soil and water is below 5%, a good correction effect can be obtained, the relative error of vegetation in the 400-600nm wave band exceeds 5%, but the maximum relative error at 400nm is still less than 10%, which indicates that the synchronous correction of vegetation also obtains a good result.
According to the ground feature verification result of the synchronous correction model, the spectrum synchronous correction method of the near-ground feature spectrometer can meet the error requirement of the spectrum, improves the range of the use condition of the ground feature spectrometer, and can carry out continuous measurement under different illuminations.
Claims (2)
1. A synchronous correction method for field measurement spectrum of a geological spectrometer is characterized by comprising the following steps: step one, establishing a fitting model of incident radiation brightness and illumination intensity
1-1, creating a data set
Detecting the incident radiance spectrum once or for many times through a standard white board and a ground object spectrometer under different illumination intensities to obtain a comprehensive data set; the obtained incident radiance spectrum comprises incident radiance values of every 1nm wavelength in a 400-900nm waveband; splitting the comprehensive data set into a training set and a test set; the Field Spec4 portable surface feature spectrometer is adopted as the surface feature spectrometer; the illuminometer is an ST-85 illuminometer;
1-2, establishing a fitting model
Fitting the data of each wave band of the incident radiance spectrum in the training set respectively to obtain the incident radiance Rad of each wave band m And the illumination intensity lux as shown in formula (1), with a wavelength m =400,401, \ 8230;, 900; the fitting process is completed by a least square method;
Rad m ＝f m (lux) formula (1)
In the formula (1), f m Fitting function for m wavelength incident radiation brightness and illumination intensity;
1-3, verifying the fitting function through the data of the test set
Respectively substituting the illumination intensity data in the test set data into fitting functions shown as the formula (1) to calculate the light of the ith sample in the test setIllumination intensity lux i ' corresponding predicted incident radiance spectral data Rad i (prediction) as shown in formula (2), i =1,2, \8230;, n; n is the number of samples in the test set;
Rad i (prediction) = (f) 400 (lux i ′)，f 401 (lux i ′)，…，f 900 (lux i ')) formula (2)
Calculating fitting accuracy ref of fitting model im The expression of (b) is shown as formula (3); ref (r) f im Representing the fitting precision of the fitting model to the ith sample of the test set under the m wave band; i =1,2, \8230;, n; m =400,401, \ 8230;, 900;
in formula (3), rad im (true) is the true incident radiance at m-band for the ith sample of the test set; rad im (prediction) predicting the incident radiance at m-band for the ith sample of the test set;
calculating the root mean square error RMSE of the fitting function m The expression of (b) is shown as formula (4);
if RMSE m If the value is higher than the preset value, the step 1-1 is executed again, and the fitting function is updated;
step two, synchronous correction of ground object reflection spectrum
Collecting the illumination intensity lux' in the environment while the ground object spectrometer collects the brightness of the ground object reflected radiation; calculating and predicting a brightness spectrum Rad (input) of incident radiation as shown in a formula (5);
rad (in) = (f) 400 (lux″)，f 401 (lux″)，…，f 900 (lux″)) Into Formula (5)
Calculating synchronous correction reflection spectrum ref of ground object Synchronization As shown in formula (6);
in the formula (6), rad (reverse) is a ground object reflection radiance spectrum of 400-900nm wave band obtained by measurement of a ground object spectrometer.
2. The method for synchronous correction of the spectrum measured in the field of a geophysical spectrometer according to claim 1, wherein the method comprises the following steps: in the step 1-1, the training set accounts for 2/3 of the total amount of data in the comprehensive data set; the test set accounts for 1/3 of the total amount of data in the integrated data set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010109585.6A CN111272664B (en) | 2020-02-22 | 2020-02-22 | Synchronous correction method for field measurement spectrum of geophysical spectrometer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010109585.6A CN111272664B (en) | 2020-02-22 | 2020-02-22 | Synchronous correction method for field measurement spectrum of geophysical spectrometer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111272664A CN111272664A (en) | 2020-06-12 |
| CN111272664B true CN111272664B (en) | 2023-03-17 |
Family
ID=71000467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010109585.6A Active CN111272664B (en) | 2020-02-22 | 2020-02-22 | Synchronous correction method for field measurement spectrum of geophysical spectrometer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111272664B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102590801A (en) * | 2012-01-18 | 2012-07-18 | 中国人民解放军61517部队 | Shadow spectrum simulating method |
| CN102901516A (en) * | 2012-09-29 | 2013-01-30 | 航天恒星科技有限公司 | Multispectral image radiation correction method based on absolute radiometric calibration |
| CN109253976A (en) * | 2018-10-22 | 2019-01-22 | 北京麦飞科技有限公司 | EO-1 hyperion real-time radiation calibrating method based on light sensation module |
| WO2020015326A1 (en) * | 2018-07-19 | 2020-01-23 | 山东科技大学 | Remote sensing image cloud shadow detection method supported by earth surface type data |
| CN110726474A (en) * | 2019-09-04 | 2020-01-24 | 南京艾格赛弗环境科技有限公司 | Hyperspectral high-frequency observation system and spectral observation method for ground objects |
-
2020
- 2020-02-22 CN CN202010109585.6A patent/CN111272664B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102590801A (en) * | 2012-01-18 | 2012-07-18 | 中国人民解放军61517部队 | Shadow spectrum simulating method |
| CN102901516A (en) * | 2012-09-29 | 2013-01-30 | 航天恒星科技有限公司 | Multispectral image radiation correction method based on absolute radiometric calibration |
| WO2020015326A1 (en) * | 2018-07-19 | 2020-01-23 | 山东科技大学 | Remote sensing image cloud shadow detection method supported by earth surface type data |
| CN109253976A (en) * | 2018-10-22 | 2019-01-22 | 北京麦飞科技有限公司 | EO-1 hyperion real-time radiation calibrating method based on light sensation module |
| CN110726474A (en) * | 2019-09-04 | 2020-01-24 | 南京艾格赛弗环境科技有限公司 | Hyperspectral high-frequency observation system and spectral observation method for ground objects |
Non-Patent Citations (2)
| Title |
|---|
| 基于野外地物光谱时间序列的反射率测量方法;董毅等;《红外》;20160110;第37卷(第01期);第31-35页 * |
| 野外地物光谱测量技术及方法;何挺等;《中国土地科学》;20021030;第16卷(第05期);第30-36页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111272664A (en) | 2020-06-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104198396B (en) | The method that polarization-hyperspectral technique diagnosis crop N P and K wanes | |
| CN108801934A (en) | A kind of modeling method of soil organic carbon EO-1 hyperion prediction model | |
| CN111855595B (en) | Spectral data calibration method based on black and white calibration plate | |
| US4779980A (en) | Atmospheric optical calibration system | |
| Rautiainen et al. | Coupling forest canopy and understory reflectance in the Arctic latitudes of Finland | |
| Liu et al. | Improving chlorophyll fluorescence retrieval using reflectance reconstruction based on principal components analysis | |
| Thome et al. | Results of the 1996 earth observing system vicarious calibration joint campaign at Lunar Lake Playa, Nevada (USA) | |
| CN109253976B (en) | High-spectrum real-time radiometric calibration method based on light sensing module | |
| CN108520127A (en) | A kind of EO-1 hyperion inversion method of seeds leaf area index | |
| CN102628800A (en) | Method and system for measuring plant leaf chlorophyll content | |
| CN105136686B (en) | The measuring method of Purple Leaf Plum anthocyanidin content | |
| CN105181606A (en) | Hyperspectral imaging technology-based method for detecting sucrose content distribution of peanut | |
| Galleano et al. | Results of the fifth international spectroradiometer comparison for improved solar spectral irradiance measurements and related impact on reference solar cell calibration | |
| CN114676636A (en) | Grassland area soil moisture rapid inversion method integrating vegetation and habitat characteristics | |
| CN111272664B (en) | Synchronous correction method for field measurement spectrum of geophysical spectrometer | |
| CN111220552B (en) | Chlorophyll hyperspectral inversion method considering illumination direction leaf radiation transmission model | |
| Julitta | Optical proximal sensing for vegetation monitoring | |
| CN114136445B (en) | Method for improving airborne hyperspectral radiation correction precision of unmanned aerial vehicle | |
| CN107945162A (en) | Detection recognition method based on computer vision technique to flourishing ant nest | |
| Bassani et al. | Laboratory activity for a new procedure of MIVIS calibration and relative validation with test data | |
| CN101813518A (en) | Method and device for measuring photometric quantity of mesopic vision | |
| CN115372310B (en) | Method and system for predicting elastic modulus of wetland by near infrared spectrum technology | |
| CN114166793B (en) | Leaf chlorophyll a and b content inversion method based on spectrum band overlapping separation | |
| Kováč et al. | Estimation of photosynthetic dynamics in forests from daily measured fluorescence and PRI data with adjustment for canopy shadow fraction | |
| Ni et al. | The spectral calibration method for a crop nitrogen sensor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |