Bionic coating spectrum consistency evaluation method adopting characteristic wave band information

Abstract

The invention discloses a bionic coating spectral consistency evaluation method adopting characteristic wave band information. The method comprises the following steps: step 1, image data preprocessing; step 2, spatial feature extraction; step 3, spectral feature extraction; step 5, combined evaluation of space-spectral feature information; and step 6, instructive improvement suggestions are given to the preparation process of the bionic coating. The evaluation method disclosed by the invention can accurately calculate and comprehensively evaluate the red edge peak wave band position, the chlorophyll response wave band, the reflectivity of the near-infrared plateau and the spectral slope of the red edge wave band, and intuitively perform joint analysis and evaluation on space and spectral characteristics, so that the evaluation result has higher reference and objectivity.

Description

Technical Field

[0001] The invention belongs to the field of information intelligent remote sensing observation, target characteristics and analysis technology, and particularly relates to a bionic coating spectrum consistency evaluation method using characteristic band information in this field. Background Art

[0002] Hyperspectral equipment can simultaneously acquire continuous images of hundreds of bands, which contain rich spatial, radiation and spectral triple information of ground objects, forming an image cube with a wide spectral range, multiple bands and high spectral resolution. With the continuous development of hyperspectral remote sensing technology, the spectral resolution and spatial resolution of imaging spectral sensors are gradually improved. It can directly perform quantitative analysis of spectral differences of camouflage materials through extremely high spectral resolution, and has great application potential in camouflage target detection.

[0003] With the continuous development of advanced coating technology, the continuous upgrading and improvement of bionic camouflage barrier preparation process can achieve high similarity with the background spectral reflectivity in specific scenarios. It has been widely used in the stealth operation field of many entities such as ships, bridges, buildings and tanks, and can effectively improve the battlefield survivability of important targets and facilities of our army. The product upgrade of bionic coatings, the improvement of modulation process, and whether it can meet the consistency of spectral reflectivity with background entities are the key to evaluating whether it meets the battlefield camouflage requirements under field conditions.

[0004] An analysis is conducted on the fields of intelligent information remote sensing observation, target characteristics and analysis technology. The spectral characteristics of traditional camouflage coatings are mostly based on indoor spectrophotometers, and spectral consistency is measured based on three influencing factors: spectral channel, spectral distance and spectral angle. However, in a complex battlefield environment, there are many influencing factors such as meteorology and lighting. The spectral reflectance curve of the bionic coating obtained by the imaging spectrometer is quite different from the spectral curve collected by the spectrophotometer under ideal indoor conditions, which will affect the evaluation results to a certain extent. In addition, how to extract more feature information from the "spectral fingerprint" to achieve a comprehensive evaluation of the high-spectral camouflage effect combining spatial and spectral characteristics, and more efficiently compare the consistency of the spectral characteristics of bionic coatings and standard green plants in a complex battlefield environment is also a topic that needs to be studied urgently.

[0005] In summary, it is particularly urgent and important to study and implement a method for evaluating the spectral consistency of bionic coatings using characteristic band information. Summary of the invention

[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art such as the single and independent spectral evaluation method of bionic coatings, the large difference between spectral data and laboratory data in complex battlefield environments, and the insufficient depth of spectral information mining, and to provide a bionic coating spectral consistency evaluation method using characteristic band information.

[0007] The present invention adopts the following technical solution:

[0008] A method for evaluating the spectral consistency of bionic coatings using characteristic band information, the improvement of which is that it comprises the following steps:

[0009] Step 1, image data preprocessing:

[0010] Based on the original hyperspectral image data, the bionic paint area and the standard green plant area are delineated to obtain the regional average spectrum; the spectral smoothing algorithm is used to obtain the smooth spectral curves of the bionic paint and the standard green plant; the spectral envelope removal algorithm is used to obtain the extreme value band of the spectral vector;

[0011] Step 2, spatial feature extraction:

[0012] Step 21: Calculate the visible light brightness contrast based on the smoothed spectral curves of the bionic paint and the standard green plants. contrast :

[0013]

[0014] In the above formula, X, Y, and Z are the three stimulus values ​​of the CIE1931 standard colorimetric system, and k is the normalization coefficient;

[0015]

[0016] S(λ) is the CIE standard illuminant D 65 The relative spectral power distribution of is the standard colorimetric observer chromaticity matching function, R(λ) is the spectral reflectance factor of the object, Δλ is the wavelength interval, and the wavelength λ ranges from 380nm to 780nm;

[0017]

[0018] In the above formula, Y target is the tristimulus value of the target color in the XYZ colorimetric system, Y background is the tristimulus value of the standard color in the XYZ colorimetric system;

[0019] Step 22: Calculate the near infrared brightness contrast NIRLightness based on the smoothed spectral curves of the bionic coating and the standard green plants contrast :

[0020] Calculate the near-infrared brightness coefficient β:

[0021]

[0022] In the above formula, E(λ) is the spectral irradiance of solar radiation outside the atmosphere, τ(λ) is the spectral transmittance of the standard atmosphere in the vertical direction from the ground to the outside of the atmosphere, K(λ) is the spectral responsivity of the sensor, and the wavelength λ ranges from 750nm to 1100nm;

[0023]

[0024] In the above formula, β target is the near-infrared brightness coefficient of the target color, β background is the near-infrared brightness coefficient of the background color;

[0025] Step 23: Calculate the color difference based on the smoothed spectral curves of the bionic paint and the standard green plants

[0026] Convert the reflectance corresponding to the smooth spectral curve to L through the XYZ color space * a * b * space:

[0027]

[0028] In the above formula, L * 、a * , b * is the coordinate value of the rectangular coordinate system, X0, Y0, and Z0 are the three stimulus values ​​of the completely diffuse reflecting surface under the illumination of the standard lighting body;

[0029] Step 3, spectral feature extraction:

[0030] Step 31, calculating the normalized difference vegetation index NDVI from the spectral imaging data of the bionic coating:

[0031]

[0032] In the above formula, NIR is the reflectivity of the near infrared band, and RED is the reflectivity of the red light band;

[0033] Step 32, calculate the spectrum red edge absorption depth evaluation index SRADI:

[0034] Calculate the peak position and peak depth of the bionic coating near 680nm, obtain the absorption valley point M near 680nm, and the two absorption shoulders S1 and S2 near 550nm and 780nm;

[0035] Calculate the spectral red edge absorption depth SRAD:

[0036]

[0037] In the above formula, the coordinates of the absorbing shoulder end S1 are (x s1 ,y s1 ), the coordinates of the absorbing shoulder S2 are (x s2 ,y s2 ), the coordinates of the absorption valley point M are (x M ,y M ), the reflectance value corresponding to the absorption valley point is reflectance M ;

[0038] Calculate the spectral red edge absorption depth slope SRADS:

[0039] SRADS=SRAD+Slope RE

[0040] In the above formula, Slope RE is the spectral slope of the red edge band calculated by the maximum value of the first-order derivative of the spectrum;

[0041] SRADI=SRADS target / SRADS background ;

[0042] In the above formula, SRADS target is the absorption depth slope of the red edge of the spectrum of the biomimetic coating, SRADS background is the absorption depth slope of the red edge of the spectrum of the background environment;

[0043] Step 4, calculate the spectral channel matching P(f):

[0044]

[0045] In the above formula, is the number of wavelength i bands whose reflectivity is not greater than the maximum reflectivity and not less than the minimum reflectivity, is the maximum value of the standard reflectivity that meets the wavelength i. is the minimum value of the standard reflectivity that the wavelength i meets, total i It is the total number of bands specified in the national standard GB / T 3979-2008, and the value is 8;

[0046] Step 5: Joint evaluation of spatial-spectral feature information:

[0047] Establish a camouflage effect evaluation model:

[0048] θ target =arg|[(exp-(σ+Lightness contrast +NIRLightness contrast)+j(P(f)*Normal(SRADI))]

[0049] In the above formula, arg|*| indicates the angle of argument. Normal(*) means to obtain normalization;

[0050] Combined with the Euler formula w = a + bi, a is the x-axis, taking the spatial eigenvalue, b is the y-axis, taking the spectral eigenvalue;

[0051] Step 6: Provide guiding suggestions for improving the preparation process of bionic coatings:

[0052] If θ target If it is greater than 45°, first take θ target The complementary angle of θ is then subtracted from 45°. target If it is less than or equal to 45°, then use 45° and θ directly. target Subtract, if the difference is greater than 15°, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants;

[0053] like If it is less than 0.75, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants.

[0054] Furthermore, in step 1, the SG spectral smoothing algorithm is used to obtain the smooth spectral curves of the bionic coating and the standard green plant; and the extreme value bands of the spectral vector near 550nm and 780nm are obtained by the spectral envelope removal algorithm.

[0055] Further, in step 2, the wavelength interval Δλ is 10 nm.

[0056] The beneficial effects of the present invention are:

[0057] The evaluation method disclosed in the present invention can accurately calculate and comprehensively evaluate the red edge peak band position, the chlorophyll response band and the reflectivity of the "near-infrared plateau", and the spectral slope of the red edge band. Finally, based on the Euler formula and the idea of ​​monotonic consistency of functions, the spatial and spectral characteristics are accurately and intuitively jointly analyzed and evaluated, making the evaluation results more referenceable and objective.

[0058] The evaluation method disclosed in the present invention realizes the joint performance evaluation of the combination of hyperspectral space and spectral features, not only from the XYZ chromaticity space and L * a * b *The color space compared the brightness contrast and color difference between the prepared bionic coating and the standard green plants, and proposed the concept of the spectral red edge absorption depth evaluation index. It also conducted a comprehensive evaluation based on indicators such as the reflectivity of the characteristic band, the red edge characteristic peak band and the spectral slope of the red edge band, and used spectral channel matching to standardize the channel value range of the standard spectrum. The evaluation model of the camouflage effect combined with the high-spectral spatial-spectral characteristics was innovatively constructed based on the complex plane idea. It can better realize the comprehensive effect evaluation from the spatial characteristics and spectral characteristics. Compared with the traditional spectral characteristic comparison evaluation, the results are more accurate and intuitive, and can provide theoretical and data support for the improvement of the bionic coating modulation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the process of the evaluation method disclosed in the present invention;

[0060] Figure 2 It is a detailed flow chart of the evaluation method disclosed in the present invention;

[0061] Figure 3 It is a schematic diagram of the positions of the absorption valley point M and the two absorption shoulder ends S1 and S2;

[0062] Figure 4 This is a schematic diagram of the principle of Euler's formula. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0064] Embodiment 1, as Figure 1 , as shown in 2, this embodiment discloses a method for evaluating the spectral consistency of bionic coatings using characteristic band information, comprising the following steps:

[0065] Step 1, image data preprocessing:

[0066] Based on the original hyperspectral image data, the bionic paint area and the standard green plant area are delineated to obtain the regional average spectrum; the SG spectrum smoothing algorithm is used to obtain the smooth spectrum curves of the bionic paint and the standard green plant; the spectrum envelope removal algorithm is used to obtain the extreme value bands of the spectrum vector near 550nm and 780nm;

[0067] Step 2, spatial feature extraction:

[0068] Step 21: Calculate the visible light brightness contrast based on the smoothed spectral curves of the bionic paint and the standard green plants. contrast :

[0069]

[0070] In the above formula, X, Y, and Z are the three stimulus values ​​of the CIE1931 standard colorimetric system, and k is the normalization coefficient;

[0071]

[0072] S(λ) is the CIE standard illuminant D 65 The relative spectral power distribution of is the standard colorimetric observer chromaticity matching function, R(λ) is the spectral reflectance factor of the object, Δλ=10nm, is the wavelength interval, and the wavelength λ ranges from 380nm to 780nm;

[0073]

[0074] In the above formula, Y target is the tristimulus value of the target color in the XYZ colorimetric system, Y background is the tristimulus value of the standard color in the XYZ colorimetric system;

[0075] Step 22: Calculate the near infrared brightness contrast NIRLightness based on the smoothed spectral curves of the bionic coating and the standard green plants contrast :

[0076] Calculate the near-infrared brightness coefficient β:

[0077]

[0078] In the above formula, E(λ) is the spectral irradiance of solar radiation outside the atmosphere, τ(λ) is the spectral transmittance of the standard atmosphere in the vertical direction from the ground to the outside of the atmosphere, K(λ) is the spectral responsivity of the sensor, and the wavelength λ ranges from 750nm to 1100nm;

[0079]

[0080] In the above formula, β target is the near-infrared brightness coefficient of the target color, β background is the near-infrared brightness coefficient of the background color;

[0081] Step 23: Calculate the color difference based on the smoothed spectral curves of the bionic paint and the standard green plants

[0082] Convert the reflectance corresponding to the smooth spectral curve to L through the XYZ color space * a * b * space:

[0083]

[0084] In the above formula, L * 、a * , b * is the coordinate value of the rectangular coordinate system, X0, Y0, and Z0 are the three stimulus values ​​of the completely diffuse reflecting surface under the illumination of the standard lighting body;

[0085] Generally speaking, when Between 0 and 1, the color difference cannot be distinguished by the naked eye. Between 1 and 2, the human eye can slightly perceive it. Between 2 and 3, the color difference between materials can be distinguished slightly, but it is not very obvious. Between 3.5 and 5, the color difference is very obvious. The color difference contrast is defined with a boundary value of 5, and the calculation formula is as follows:

[0086]

[0087] Step 3, spectral feature extraction:

[0088] Step 31, calculating the normalized difference vegetation index NDVI from the spectral imaging data of the bionic coating:

[0089]

[0090] In the above formula, NIR is the reflectivity of the near infrared band, and RED is the reflectivity of the red light band;

[0091] See the difference in reflectance between biomimetic paint and standard greenery in the near-infrared and red wavelengths.

[0092] Step 32, calculate the spectrum red edge absorption depth evaluation index SRADI:

[0093] Calculate the peak position and peak depth of the bionic coating near 680nm, such as Figure 3 As shown, an absorption valley point M near 680nm and two absorption shoulders S1 and S2 near 550nm and 780nm are obtained;

[0094] Calculate the spectral red edge absorption depth SRAD (Spectral REP Absorb Depth):

[0095]

[0096] In the above formula, the coordinates of the absorbing shoulder end S1 are (x s1 ,y s1 ), the coordinates of the absorbing shoulder S2 are (x s2 ,y s2 ), the coordinates of the absorption valley point M are (x M ,y M), the reflectance value corresponding to the absorption valley point is reflectance M ;

[0097] Calculate the spectral red edge absorption depth slope SRADS (Spectral REP Absorb DepthSlope):

[0098] SRADS=SRAD+Slope RE

[0099] In the above formula, Slope RE is the spectral slope of the red edge band calculated by the maximum value of the first-order derivative of the spectrum;

[0100] SRADI=SRADS target / SRADS background ;

[0101] In the above formula, SRADS target is the absorption depth slope of the red edge of the spectrum of the biomimetic coating, SRADS background is the absorption depth slope of the red edge of the spectrum of the background environment;

[0102] Step 4: Calculate the spectral channel matching P(f) using the probability addition method:

[0103]

[0104] In the above formula, is the number of wavelength i bands whose reflectivity is not greater than the maximum reflectivity and not less than the minimum reflectivity, is the maximum value of the standard reflectivity that meets the wavelength i. is the minimum value of the standard reflectivity that the wavelength i meets, total i It is the total number of bands specified in the national standard GB / T 3979-2008, and the value is 8;

[0105] Step 5: Joint evaluation of spatial-spectral feature information:

[0106] Establish a camouflage effect evaluation model:

[0107] θ target =arg|[(exp-(σ+Lightness contrast +NIRLightness contrast )+j(P(f)*Normal(SRADI))]

[0108] In the above formula, arg|*| indicates the angle of argument. Normal(*) means to obtain normalization;

[0109] like Figure 4 As shown, combined with the Euler formula w = a + bi = rcosθ + jrsinθ, a is the x-axis, taking the spatial eigenvalue, b is the y-axis, taking the spectral eigenvalue;

[0110] Step 6: Provide guiding suggestions for improving the preparation process of bionic coatings:

[0111] If θ target If it is greater than 45°, first take θ target The complementary angle of θ is then subtracted from 45°. target If it is less than or equal to 45°, then use 45° and θ directly. target Subtract, if the difference is greater than 15°, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants;

[0112] like If it is less than 0.75, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants.

Claims

1. A method for evaluating the spectral consistency of bionic coatings using characteristic band information, characterized in that: The steps include: Step 1, image data preprocessing: Based on the original hyperspectral image data, the bionic paint area and the standard green plant area are delineated to obtain the regional average spectrum; the spectral smoothing algorithm is used to obtain the smooth spectral curves of the bionic paint and the standard green plant; the spectral envelope removal algorithm is used to obtain the extreme value band of the spectral vector; Step 2, spatial feature extraction: Step 21: Calculate the visible light brightness contrast based on the smoothed spectral curves of the bionic paint and the standard green plants. contrast : In the above formula, X, Y, and Z are the three stimulus values ​​of the CIE1931 standard colorimetric system, and k is the normalization coefficient; S(λ) is the CIE standard illuminant D 65 The relative spectral power distribution of is the standard colorimetric observer chromaticity matching function, R(λ) is the spectral reflectance factor of the object, Δλ is the wavelength interval, and the wavelength λ ranges from 380nm to 780nm; In the above formula, Y target is the tristimulus value of the target color in the XYZ colorimetric system, Y background is the tristimulus value of the standard color in the XYZ colorimetric system; Step 22: Calculate the near infrared brightness contrast NIRLightness based on the smoothed spectral curves of the bionic coating and the standard green plants contrast : Calculate the near-infrared brightness coefficient β: In the above formula, E(λ) is the spectral irradiance of solar radiation outside the atmosphere, τ(λ) is the spectral transmittance of the standard atmosphere in the vertical direction from the ground to the outside of the atmosphere, K(λ) is the spectral responsivity of the sensor, and the wavelength λ ranges from 750nm to 1100nm; In the above formula, β target is the near-infrared brightness coefficient of the target color, β background is the near-infrared brightness coefficient of the background color; Step 23: Calculate the color difference based on the smoothed spectral curves of the bionic paint and the standard green plants Convert the reflectance corresponding to the smooth spectral curve to L through the XYZ color space * a * b * space: In the above formula, L * 、a * 、b * is the coordinate value of the rectangular coordinate system, X0, Y0, and Z0 are the three stimulus values ​​of the completely diffuse reflecting surface under the illumination of the standard lighting body; Step 3, spectral feature extraction: Step 31, calculating the normalized difference vegetation index NDVI from the spectral imaging data of the bionic coating: In the above formula, NIR is the reflectivity of the near infrared band, and RED is the reflectivity of the red light band; Step 32, calculate the spectrum red edge absorption depth evaluation index SRADI: Calculate the peak position and peak depth of the bionic coating near 680nm, obtain the absorption valley point M near 680nm, and the two absorption shoulders S1 and S2 near 550nm and 780nm; Calculate the spectral red edge absorption depth SRAD: In the above formula, the coordinates of the absorbing shoulder end S1 are (x s1 ,y s1 ), the coordinates of the absorbing shoulder S2 are (x s2 ,y s2 ), the coordinates of the absorption valley point M are (x M ,y M ), the reflectance value corresponding to the absorption valley point is reflectance M ; Calculate the spectral red edge absorption depth slope SRADS: SRADS=SRAD+Slope RE In the above formula, Slope RE is the spectral slope of the red edge band calculated by the maximum value of the first-order derivative of the spectrum; SRADI=SRADS target / SRADS background , In the above formula, SRADS target is the absorption depth slope of the red edge of the spectrum of the biomimetic coating, SRADS background is the absorption depth slope of the red edge of the spectrum of the background environment; Step 4, calculate the spectral channel matching P(f): In the above formula, is the number of wavelength i bands whose reflectivity is not greater than the maximum reflectivity and not less than the minimum reflectivity, is the maximum value of the standard reflectivity that meets the wavelength i. is the minimum value of the standard reflectivity that the wavelength i meets, total i It is the total number of bands specified in the national standard GB / T 3979-2008, and the value is 8; Step 5: Joint evaluation of spatial-spectral feature information: Establish a camouflage effect evaluation model: θ target =arg|[(exp-(σ+Lightness contrast +NIRLightness contrast )+j(P(f)*Normal(SRADI))] In the above formula, arg|*| indicates the angle of argument. Normal(*) means to obtain normalization; Combined with the Euler formula w = a + bi, a is the x-axis, taking the spatial eigenvalue, b is the y-axis, taking the spectral eigenvalue; Step 6: Provide guiding suggestions for improving the preparation process of bionic coatings: If θ target If it is greater than 45°, first take θ target The complementary angle of θ is then subtracted from 45°. target If it is less than or equal to 45°, then use 45° and θ directly. target Subtract, if the difference is greater than 15°, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants; like If it is less than 0.75, it is necessary to enhance the consistency of the spatial spectrum characteristics between the bionic coating and the standard green plants.

2. The method for evaluating the spectral consistency of bionic coatings using characteristic band information according to claim 1, characterized in that: In step 1, the SG spectral smoothing algorithm is used to obtain the smooth spectral curves of the bionic coating and the standard green plant; the extreme value bands of the spectral vector near 550nm and 780nm are obtained by the spectral envelope removal algorithm.

3. The method for evaluating the spectral consistency of bionic coatings using characteristic band information according to claim 1, characterized in that: In step 2, the wavelength interval Δλ is 10 nm.

Images