Underwater target spectral reflectivity reconstruction method based on differential calculation

Abstract

The invention discloses an underwater target spectral reflectivity reconstruction method based on differential calculation, and solves the technical problem that the existing underwater object spectral reflectivity acquisition method is inaccurate in measurement or is only suitable for short-distance shooting and cannot be used for calculating the spectral reflectivity of a long-distance object. According to the method, extra distance measuring equipment is not needed, only one polarization spectral imager is used for mobile shooting to obtain two groups of images similar to binocular imaging, and the distance and angle information of the light source, the imager and the target point is calculated through a binocular measurement algorithm and a solid geometry model. And meanwhile, the original incident and emergent radiance value of the target object can be calculated by combining a water body attenuation coefficient measured in situ by an attenuation instrument and a Lambert-Beer law, so that the real spectral reflectivity of the target is reconstructed.

Description

Technical Field

[0001] The invention relates to an underwater spectral reflectivity measurement method, and in particular to an underwater target spectral reflectivity reconstruction method based on differential calculation. Background Art

[0002] With the development of marine technology, spectral imagers have the ability to observe targets in deep waters. Many researchers use manned or unmanned underwater vehicles equipped with spectral imagers to collect underwater spectral information on a large scale. The spectral reflectance data measured can be used to investigate underwater ecological environment, monitor the growth and health of seabed plants and animals, or evaluate the distribution and abundance of underwater mineral resources or fishery resources. Therefore, obtaining effective and accurate spectral reflectance of underwater targets can greatly promote the detection of other marine research fields.

[0003] However, there are few studies on the measurement of spectral reflectance of underwater objects, because light propagating in the air is less affected by atmospheric absorption and scattering, the attenuation is low, and the light source characteristics are stable. The spectral reflectance can be obtained by simply using a radiometer to measure the outgoing radiance value of the standard reflectance plate and the outgoing radiance value of the object, and performing a ratio operation on the two. However, due to the complex underwater imaging environment, the spectral imager is affected by the ambient light from the sun, the active light source, the scattered light emitted by the suspended particles in the water absorbing the ambient light and the active light, and the attenuation of the light emitted by the target object due to scattering and absorption during propagation. In addition, the water body has different absorption degrees for light of different wavelengths, which makes the spectral reflectance of the object directly measured underwater different from that in the air. Therefore, the reflectance measurement method in the air cannot be directly applied to underwater.

[0004] In current research, there are two main methods for obtaining the spectral reflectance of underwater objects. One is that the Norwegian University of Science and Technology places a calibration plate with known reflectance in the image captured by the spectral imager, and performs a ratio operation on the radiance value of each pixel in the image and the outgoing radiance value of the calibration plate. This is the same as the method of measuring the spectral reflectance of objects in the air, but this method has many defects: the light field distribution in the water body is uneven, the incident light intensity received by each pixel in the image is different, and the distance between each pixel and the calibration plate and the imager is different, which leads to inaccurate measurement of the spectral reflectance. Another method is to establish an underwater imaging model, calculate and estimate each component in the model through measuring instruments and algorithms, and finally obtain the true spectral reflectance of the object. Zhejiang University uses binocular spectral imagers, active light sources and water attenuation meters to measure the light field distribution of the light source in the calculation model, the distance information between the target point and the imager and the light source, and the water attenuation coefficient to obtain the spectral reflectance of underwater objects, but it ignores the influence of the scattered light of suspended particles on the image captured by the spectral imager. It is only suitable for close-range shooting and cannot be used to calculate the spectral reflectance of distant objects. Summary of the invention

[0005] The purpose of the present invention is to solve the technical problems that the existing methods for obtaining the spectral reflectance of underwater objects have inaccurate measurements or are only suitable for close-range shooting and cannot be used for calculating the spectral reflectance of distant objects, and to provide a method for reconstructing the spectral reflectance of underwater targets based on differential calculation.

[0006] The concept of the present invention is: using a polarization spectrum imager to take images of underwater targets when the light source is turned on and off, and performing differential removal of the influence of the solar ambient light. Then, polarization information and Stokes vector differential removal of the influence of backscattered light are used to obtain target information light. By moving the polarization spectrum imager and the light source as a whole to take adjacent frame images for calculation, the distance and angle information between the target point and the spectrometer and the light source in the target scene are obtained, and then the known target information light, distance angle information, light source irradiance information and water body attenuation coefficient are used to reconstruct the real spectral reflectance of the underwater target.

[0007] In order to achieve the above invention objectives and complete the above invention concepts, the present invention adopts the following technical solutions:

[0008] A method for reconstructing the spectral reflectance of an underwater target based on differential calculation is special in that it includes the following steps:

[0009] S1. Perform radiation calibration on the polarization spectroscopic imager to obtain the radiation calibration coefficient C(λ), and use the Zhang Zhengyou calibration method to calculate the internal parameter matrix and distortion coefficient of the polarization spectroscopic imager. At the same time, use a standard spectral radiometer to measure the irradiance E at 1 meter away from the light source. 0 (λ);

[0010] S2. Place the polarization spectrum imager and the light source at the same height underwater, with a distance of l between them. camera-light , and are in the same spatial plane, turn on the light source to illuminate the target scene, and use the water body attenuation coefficient measuring instrument to measure the current water body attenuation coefficient α(λ);

[0011] S3, using a polarization spectral imager to shoot the target scene when the light source is turned on and off, respectively, to obtain two sets of multispectral image sequences, where the multispectral image sequence when the light source is turned on is The multispectral image sequence when the light source is turned off is

[0012] S4, move the polarization spectrum imager and the light source horizontally for a certain distance as a whole, repeat S3, and obtain two sets of multispectral image sequences, where the multispectral image sequence when the light source is turned on is The multispectral image sequence when the light source is turned off is

[0013] S5, through the radiation calibration coefficient, the two sets of multispectral image sequences in S3 are and S4 two sets of multispectral image sequences Converted into radiance values ​​before moving and the radiance value after moving The target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source is calculated, and the spectrum images after removing the backscattered light before and after the overall movement are obtained respectively;

[0014] S6. Select spectral images of specific bands from the spectral images after removing backscattered light before and after the overall movement according to the turbidity of the water body, pre-process the two spectral images in combination with the internal parameter matrix and distortion coefficient of the polarization spectral imager, and determine the overlapping area of ​​the two images, and calculate the distance of each pixel point in the overlapping area corresponding to the target object scene from the optical center of the polarization spectral imager and the light source before the movement and the incident angle of the light source in combination with binocular ranging;

[0015] S7, based on the current water body attenuation coefficient α(λ), the irradiance E at 1 meter from the light source 0 The spectral reflectance of the target scene is reconstructed based on the target information light radiance value (λ) before the overall movement, the distance between the target point in the target scene corresponding to each pixel point in the overlapping area and the optical center of the polarization spectrum imager and the light source, and the incident angle of the light source, and the spectral reflectance reconstruction of the underwater target is completed.

[0016] Furthermore, in S1, the polarization mode of the polarization spectrum imager is linear polarization, and the polarization angles are 0°, 45°, 90° and 135°.

[0017] Furthermore, the specific process of S5 is as follows:

[0018] S5.1, take S3 two groups of multispectral image sequences as: turn on the light source, the multispectral image sequence is Turn off the light source and the multispectral image sequence is S4 Two sets of multispectral image sequences are represented as follows: Turn on the light source, and the multispectral image sequence is Turn off the light source and the multispectral image sequence is Among them, x, y represent the horizontal and vertical coordinates of the spectral image sequence, λ represents the incident wavelength, represents the polarization angle;

[0019] S5.2, the four multispectral image sequences in S5.1 are converted into radiance values ​​by using the radiation calibration coefficients, which are expressed as: and

[0020] S5.3. Subtract the multi-spectral image sequences of the light source turned on and off at the same position respectively to obtain the spectral image sequence of the active light source at that position; remove the backscattered light of the spectral image sequence of the active light source, and calculate the target information light radiance value of the target object scene entering the imaging spectrometer before and after the overall movement of the polarization spectral imager and the light source respectively, to obtain the spectral image after removing the backscattered light before and after the overall movement.

[0021] Further, in S5.2, the formula for converting the four multispectral image sequences in S5.1 into radiance values ​​by using the radiation calibration coefficient is:

[0022]

[0023] Among them, x is one of camera-light-L, camera-L, camera-light-R, camera-R; C(λ) represents the radiation calibration coefficient.

[0024] Further, in S5.3, the multi-spectral image sequences of the light source turned on and turned off at the same position are subtracted to obtain the spectral image sequence of the active light source at the position, specifically:

[0025] The multi-spectral image sequence of the light source turned on and off at the position before the polarization spectral imager and the light source move as a whole is subtracted to obtain the active light source spectral image sequence L at that position. light-L (x,y,λ) is:

[0026] L light-L (x,y,λ)=L camera-light-L (x,y,λ)-L camera-L (x,y,λ)

[0027] After the polarization spectral imager and the light source are moved as a whole, the multi-spectral image sequences of the light source turned on and off at the position are subtracted to obtain the active light source spectral image sequence L at that position. light-R (x,y,λ) is:

[0028] L light-R (x,y,λ)=L camera-light-R (x,y,λ)-L camera-R (x,y,λ).

[0029] Further, in S5.3, the target information light radiance values ​​of the target scene entering the imaging spectrometer before and after the polarization spectrum imager and the light source move as a whole are calculated as follows:

[0030] The Stokes vector representation of the target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source is calculated as follows:

[0031]

[0032] Among them, S m In (x, y, λ), m is light-L and light-R, which respectively represent the target information light radiance value of the target scene entering the imaging spectrometer before the polarization spectrum imager and the light source move as a whole, and the target information light radiance value of the target scene entering the imaging spectrometer after the polarization spectrum imager and the light source move as a whole; p scat (λ) represents the polarization degree of the backscattered light of the active light source spectrum image sequence, represents the optimal target information light polarization degree p in the active light source spectrum image sequence obj (λ);

[0033] I m (x,y,λ)=L m (x,y,λ,90°)+L m (x,y,λ,0°);

[0034] Q m (x,y,λ)=L m (x,y,λ,90°)-L m (x,y,λ,0°);

[0035] U m (x,y,λ)=L m (x,y,λ,45°)-L m (x,y,λ,135°);

[0036] in, Represents the active light source spectrum image sequence L m (x,y,λ) The image obtained by polarization of the angle;

[0037] The calculation process of the polarization degree of backscattered light of the active light source spectrum image sequence is:

[0038] In the active light source spectrum image sequence corresponding to each wavelength, a target-free area σ is framed. void (λ), the degree of polarization of backscattered light of pixels in the active light source spectrum image sequence is estimated as:

[0039]

[0040] B max (x,y,λ) is σ void The maximum backscattered light radiance in (λ), B min (x,y,λ) is σ void (λ) is the minimum backscattered light radiance; n is σ voidThe number of pixels in (λ);

[0041] The polarization degree p of the target information light in the active light source spectrum image sequence is obj (λ) takes values ​​from 0 to 1 in steps of 0.01, and frames the region containing a clear single target in the active light source spectrum image sequence at each wavelength. clear (λ), the optimal polarization value of the target information light is calculated using the mutual information of the target information light radiance value and the backscattered light radiance value. The formula is:

[0042]

[0043] In the formula, MI[B,S] represents B m (x,y,λ) and S m The mutual information value of (x, y, λ), B m (x, y, λ) represents the backscattered light radiance value, represents the optimal p obj (λ).

[0044] Furthermore, S6 is specifically:

[0045] S6.1. Under a certain water turbidity, select two multispectral image sequences with the highest contrast in the spectrum before and after the overall movement after removing the backscattered light, and use the spectrum image before the movement as the left image; and use the spectrum image after the movement as the right image;

[0046] S6.2, using the intrinsic parameter matrix and distortion coefficient of the polarization spectral imager and the pose information of the polarization spectral imager when shooting the target scene, perform epipolar correction on the left image and the right image after removing the backscattered light, perform stereo matching of the two images by using the SIFT operator, and determine the overlapping area of ​​the two spectral images;

[0047] S6.3. Calculate the depth Z of the target point P in the target scene corresponding to each pixel in the overlapping area p ;

[0048] S6.4, perform binocular visual ranging on the left image and right image after removing the backscattered light, and combine the depth Z p , the three-dimensional geometric model calculates the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the optical center of the polarization spectrum imager before the overall movement camera (x, y), the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the light source before the overall movement light (x,y) and the incident angle of the light source.

[0049] Furthermore, S6.3 is specifically:

[0050] The optical center of the polarization spectrum imager before moving is taken as the origin O Cl , take the optical center of the mobile polarization spectrum imager as the origin O Cr , establish two coordinate systems with the moving direction as the x-axis, the vertical upward direction as the z-axis, and any direction perpendicular to the x-axis and z-axis as the y-axis;

[0051] The horizontal moving distance l of the polarization spectrum imager is known L-R The horizontal coordinate values ​​of the target point P in the target scene in the overlapping area of ​​the left image and the right image are x l 、x r , the focal length f of the polarization spectrum imager, calculate the depth Z of the target point P along the z-axis P for:

[0052]

[0053] Further, in S6.4, the incident angle of the light source is the angle between the line between the light source and the target point P in the target scene before the overall movement and the target point P in the target scene at x0 Cl The angle formed by the line connecting the projections on the y plane.

[0054] Furthermore, the specific process of S7 is as follows:

[0055] S7.1, according to the Lambert-Beer law:

[0056]

[0057] Among them, S light-L (x, y, λ) represents the target information light radiance value before the overall movement, L reflect (x, y, λ) represents the radiance reflected by the target in the overlapping area under the active light before the overall movement, α(λ) represents the current water attenuation coefficient, l camera (x, y) represents the distance between the target point on the target object and the optical center of the polarization spectrum imager corresponding to each pixel in the overlapping area;

[0058] S7.2. When the light source is irradiated to each point on the spherical surface at a distance of 1m, the irradiance received by each point is the same, and the irradiance received by the distant target is related to the square of the irradiation distance. Combining the Lambert cosine theorem and the Lambert-Beer law, the irradiance E of the target scene in the overlapping area irradiated by the light source can be obtained. incidence (x,y,λ) is:

[0059]

[0060] Among them, E 0 (λ) represents the irradiance at 1 meter from the light source, θ i(x,y) represents the incident angle of the light source; l light (x, y) represents the distance between the target point and the light source on the target object corresponding to each pixel in the overlapping area;

[0061] S7.3. Calculate the spectral reflectance of the reconstructed target scene using the formula:

[0062]

[0063] Beneficial effects of the present invention:

[0064] 1. The present invention discloses a method for reconstructing the spectral reflectance of underwater targets based on differential calculation. No additional ranging equipment is required. Only one polarization spectral imager is used for mobile shooting to obtain two sets of images similar to binocular imaging. The distance and angle information of the light source, the imager and the target point are calculated through a binocular measurement algorithm and a stereo geometric model. At the same time, the original incident and outgoing radiance values ​​of the target object can be calculated by combining the water body attenuation coefficient measured in situ by an attenuator and the Lambert-Beer law, thereby reconstructing the real spectral reflectance of the target.

[0065] 2. The present invention provides a method for reconstructing the spectral reflectance of underwater targets based on differential calculation. The method utilizes an active light source flicker differential method to remove the reflected light portion of the target information light received by the spectrometer due to sunlight exposure, thereby ensuring that the target object is only illuminated by the active light source in the reflectance calculation without interference from other light sources.

[0066] 3. The present invention provides a method for reconstructing the spectral reflectance of underwater targets based on differential calculation, which uses image polarization information and Stokes vector differential to remove the influence of backscattered light. Backscattered light is the main reason for reducing image quality and interfering with the spectral characteristics of the object itself. By removing backscattered light through polarization information and obtaining object target information light, the accuracy of target reflectance reconstruction is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a flow chart of an embodiment of a method for reconstructing spectral reflectance of an underwater target based on differential calculation according to the present invention;

[0068] Figure 2 This is a schematic diagram of binocular ranging in Example S6.5 of a method for reconstructing spectral reflectance of an underwater target based on differential calculation according to the present invention;

[0069] Figure 3 It is a schematic diagram of a three-dimensional geometric model in Example S6.5 of a method for reconstructing spectral reflectance of an underwater target based on differential calculation of the present invention;

[0070] Figure 4 A light source radiation distribution diagram in an embodiment of a method for reconstructing spectral reflectance of an underwater target based on differential calculation according to the present invention;

[0071] Figure 5 It is a schematic diagram of reflectivity reconstruction in an embodiment of a method for reconstructing spectral reflectivity of an underwater target based on differential calculation in the present invention. DETAILED DESCRIPTION

[0072] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] The present invention provides a method for reconstructing the spectral reflectance of an underwater target based on differential calculation, such as Figure 1 As shown, the following steps are included:

[0074] S1. Use the spectral imager radiation calibration device (standard integrating sphere light source system, spectral radiation brightness standard lamp system, standard spectral radiometer) on water to calibrate the polarization spectral imager to obtain the pixel output value of the spectrometer. and spectral radiance The radiation calibration coefficient C(λ) is is the linear polarization direction angle, which are 0°, 45°, 90° and 135° respectively. The Zhang Zhengyou calibration method is used to analyze the internal parameter matrix and distortion coefficient of the polarization spectrum imager. At the same time, a standard spectral radiometer is used to measure the irradiance E at 1 meter away from the light source. 0 (λ).

[0075] S2. Place the polarization spectrum imager and the light source at the same height underwater, with a distance of l between them. camera-light , and are in the same spatial plane, turn on the light source to illuminate the target scene, and use the water attenuation coefficient measuring instrument to measure the current water attenuation coefficient α(λ).

[0076] S3. Use a polarization spectral imager to shoot the target scene to obtain a set of multispectral image sequences Quickly turn off the light source and use the imaging spectrometer to photograph the target again to obtain a set of multispectral image sequences.

[0077] S4, mobile polarization spectrum imager and light source overall l L-R Repeat S3 under the conditions of unchanged distance, light source and shooting angle to obtain two sets of multispectral image sequences

[0078] S5, converting the two sets of multispectral image sequences of S3 and the two sets of multispectral image sequences of S4 into radiance values ​​before and after the movement respectively through the radiation calibration coefficient, and calculating the radiance value of the target information light of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source, and obtaining the spectral images after removing the backscattered light before and after the overall movement respectively; the specific process is:

[0079] S5.1, take S3 two groups of multispectral image sequences as: turn on the light source, the multispectral image sequence is Turn off the light source and the multispectral image sequence is S4 Two sets of multispectral image sequences are represented as follows: Turn on the light source, and the multispectral image sequence is Turn off the light source and the multispectral image sequence is Among them, x, y represent the horizontal and vertical coordinates of the spectral image sequence, λ represents the incident wavelength, represents the polarization angle;

[0080] S5.2, the four multispectral image sequences in S51 are converted into radiance values ​​through the radiation calibration coefficients, which are expressed as: and

[0081] The formula for converting the four multispectral image sequences in S5.1 into radiance values ​​through the radiation calibration coefficient is:

[0082]

[0083] Among them, x is one of camera-light-L, camera-L, camera-light-R, camera-R; C(λ) represents the radiation calibration coefficient.

[0084] The radiance values ​​of the spectral images when the light source is turned on before and after the device is moved can be expressed as:

[0085]

[0086] The radiance values ​​of the spectral images with the light source turned off before and after the device is moved can be expressed as:

[0087]

[0088] Among them, S sun (x, y, λ) represents the radiance value of the target information light reflected to the imager when the object is illuminated by the external ambient light, B sun (x, y, λ) represents the radiance value of the backscattered light caused by the external ambient light.

[0089] S5.3, respectively subtract the multi-spectral image sequences of the light source turned on and turned off at the same position to obtain the spectral image sequence of the active light source at the position; specifically:

[0090] The multi-spectral image sequence of the light source turned on and off at the position before the polarization spectral imager and the light source move as a whole is subtracted to obtain the active light source spectral image sequence L at that position. light-L (x,y,λ) is:

[0091]

[0092] The multi-spectral image sequence of the light source turned on and off at the position before the polarization spectral imager and the light source move as a whole is subtracted to obtain the active light source spectral image sequence L at that position. light-R (x,y,λ) is:

[0093]

[0094] Among them, S light-L (x,y,λ),S light-R (x, y, λ) represent the target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source; B light-L (x,y,λ),B light-R (x, y, λ) represent the backscattered light radiance values ​​before and after the overall movement.

[0095] The active light source spectrum image sequence can be expressed as:

[0096] L(x,y,λ)=L max (x,y,λ)+L min (x,y,λ)(6)

[0097] Among them, L max (x,y,λ) and L min (x, y, λ) are the imaging results when the backscattered light radiance value is the maximum and minimum during the imaging process, which can be expressed as:

[0098] L max (x,y,λ)=S max (x,y,λ)+B max (x,y,λ)(7)

[0099] L min (x,y,λ)=S min (x,y,λ)+B min (x,y,λ)(8)

[0100] The degree of polarization p is expressed by the Stokes vector as

[0101]

[0102] in:

[0103]

[0104] L max (x,y,λ)-L min (x,y,λ)=p scat (λ)B(x,y,λ)+p obj (λ)S(x,y,λ)(13)

[0105] The Stokes vector representation of the target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source is calculated as follows:

[0106]

[0107] Among them, S m In (x, y, λ), m is light-L and light-R, which respectively represent the target information light radiance value of the target scene entering the imaging spectrometer before the polarization spectrum imager and the light source move as a whole, and the target information light radiance value of the target scene entering the imaging spectrometer after the polarization spectrum imager and the light source move as a whole; p scat (λ) represents the polarization degree of the backscattered light of the active light source spectrum image sequence, represents the optimal target information light polarization degree p in the active light source spectrum image sequence obj (λ);

[0108] I m (x,y,λ)=L m (x,y,λ,90°)+L m (x,y,λ,0°);

[0109] Q m (x,y,λ)=L m (x,y,λ,90°)-L m (x,y,λ,0°);

[0110] U m (x,y,λ)=L m (x,y,λ,45°)-L m (x,y,λ,135°);

[0111] in, Represents the active light source spectrum image sequence L m (x,y,λ) The image obtained by polarization of the angle;

[0112] In the active light source spectrum image sequence corresponding to each wavelength, a target-free area σ is framed. void (λ), each pixel in its area has L max (x,y,λ)≈B max (x,y,λ),L min (x,y,λ)≈B min (x,y,λ), the global backscattered light polarization degree is estimated as:

[0113]

[0114] B max (x,y,λ) is σ void The maximum backscattered light radiance in (λ), B min (x,y,λ) is σ void (λ) is the minimum backscattered light radiance; n is σ void The number of pixels in (λ);

[0115] The mutual information of the target information light and the backscattered light is

[0116]

[0117] The polarization degree p of the target information light in the active light source spectrum image sequence is obj (λ) takes values ​​from 0 to 1 in steps of 0.01, and frames the region containing a clear single target in the active light source spectrum image sequence at each wavelength. clear (λ), the optimal polarization value of the target information light is calculated using the mutual information of the target information light radiance value and the backscattered light radiance value. The formula is:

[0118]

[0119] In the formula, MI[B,S] represents B m (x,y,λ) and S m The mutual information value of (x, y, λ), B m (x, y, λ) represents the backscattered light radiance value, represents the optimal p obj (λ).

[0120] According to the above formula, the target information light radiance value S of the target scene entering the imaging spectrometer before and after the overall movement is calculated. light-L (x,y,λ) and S light-R (x,y,λ).

[0121] According to the target information light radiance value S of the target scene entering the imaging spectrometer before and after the overall movement light-L (x,y,λ) and Slight-R (x, y, λ) obtains the spectral image after removing the backscattered light before and after the overall movement.

[0122] S6. Select a spectral image of a specific band from the spectral images after removing the backscattered light before and after the overall movement according to the turbidity of the water body, pre-process the two spectral images in combination with the internal parameter matrix and distortion coefficient of the polarization spectral imager, and determine the overlapping area of ​​the two images. Calculate the distance of each pixel point in the overlapping area corresponding to the target object scene from the optical center of the polarization spectral imager and the light source before the movement and the incident angle of the light source in combination with binocular ranging; specifically:

[0123] S6.1. Under a certain water turbidity, select two multispectral image sequences with the highest contrast in the spectrum before and after the overall movement after removing the backscattered light, and use the spectrum image before the movement as the left image; and use the spectrum image after the movement as the right image;

[0124] S6.2, using the intrinsic parameter matrix and distortion coefficient of the polarization spectral imager obtained in S1 and the position information of the polarization spectral imager when shooting the target scene, perform epipolar correction on the left image and the right image after removing the backscattered light, perform stereo matching of the two images by using the SIFT operator, and determine the overlapping area of ​​the two spectral images;

[0125] S6.3. Calculate the depth Z of the target point P in the target scene corresponding to each pixel in the overlapping area p ;

[0126] The optical center of the polarization spectrum imager before moving is taken as the origin O Cl , take the optical center of the mobile polarization spectrum imager as the origin O Cr , establish two coordinate systems with the moving direction as the x-axis, the vertical upward direction as the z-axis, and any direction perpendicular to the x-axis and z-axis as the y-axis;

[0127] The horizontal moving distance l of the polarization spectrum imager is known L-R The horizontal coordinate values ​​of the target point P in the target scene in the overlapping area of ​​the left image and the right image are x l 、x r , the focal length f of the polarization spectrum imager is given by Figure 2 and the known distance l between the two imaging optical centers L-R , the horizontal coordinate value x of the target point in the left image and the right image l x r , the focal length f of the imaging spectrometer, and the following formula (19) can be used to calculate the depth Z P :

[0128]

[0129] S6.4, perform binocular visual ranging on the left image and right image after removing the backscattered light, and combine the depth Z p , the three-dimensional geometric model calculates the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the optical center of the polarization spectrum imager before the overall movement camera (x, y), the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the light source before the overall movement light (x, y) and the incident angle of the light source: Figure 2 Two pairs of similar triangles ΔO C AH~ΔO C OC and ΔO C HP~ΔO C Cp can be obtained as follows

[0130]

[0131] The coordinates of the target point P in the camera coordinate system can be calculated by equations (19) and (20), and then the distance between the target point P and the optical center O of the imaging spectrometer can be calculated: C Distance camera and distance from light source O L Distance light for

[0132]

[0133] At the same time, the incident angle θ between the incident light from the light source to the target point P and the plane where the target point is located can be calculated i for:

[0134]

[0135] According to equations (19)(20)(21)(22)(23), the distance l from each pixel in the overlapping area to the optical center of the imaging spectrometer and the light source can be calculated: camera (x,y),l light (x, y), and the incident angle θ of the light source i (x,y).

[0136] S7, based on the current water body attenuation coefficient, the irradiance at 1 meter from the light source, the target information light radiance value before the overall movement, the distance between the target point in the target scene corresponding to each pixel in the overlapping area and the optical center of the polarization spectrum imager and the light source, and the incident angle of the light source, reconstruct the spectral reflectance of the target scene and complete the reconstruction of the spectral reflectance of the underwater target. The specific process is:

[0137] S7.1, according to the Lambert-Beer law:

[0138]

[0139] Among them, S light-L (x, y, λ) represents the target information light radiance value before the overall movement, L reflect (x, y, λ) represents the radiance reflected by the target in the overlapping area under the active light before the overall movement, α(λ) represents the current water attenuation coefficient, l camera (x, y) represents the distance between the target point on the target object and the optical center of the polarization spectrum imager corresponding to each pixel in the overlapping area;

[0140] Then, the radiance reflected by the target scene in the overlapping area under active light illumination can be obtained as:

[0141]

[0142] S7.2. The irradiance of each point on the spherical surface at a distance of 1m is the same, and the irradiance of the distant target is related to the square of the irradiation distance. Combining the Lambert cosine theorem and the Lambert-Beer law, the irradiance E of the target scene in the overlapping area irradiated by the light source can be obtained. incidence (x,y,λ) is:

[0143]

[0144] Among them, E 0 (λ) represents the irradiance at 1 meter from the light source, θ i (x,y) represents the incident angle of the light source; l light (x, y) represents the distance between the target point and the light source on the target object corresponding to each pixel in the overlapping area;

[0145] S7.3. Calculate the spectral reflectance of the reconstructed target scene using the formula:

[0146]

[0147] The invention discloses a method for reconstructing the spectral reflectivity of an underwater target based on differential calculation. The method uses polarization information and a Stokes vector to remove the influence of backscattered light in an imaging process, uses a method for differencing adjacent frame images to calculate the distance between a device (imager, light source) and all target points in an image, and uses a method for light field difference to remove the influence of external ambient light on data inversion, thereby avoiding defects in existing reflectivity calculation methods (in the process of inverting the real reflectivity data of an object, the distance between the object and the imager must be the same as the distance between the object and the imager and the backscattered light caused by the active light source for long-distance imaging cannot be ignored). The underwater imaging model and the Lambert-Beer law are used to accurately calculate the incident light radiance value of the active light source and the radiance value of the received information light of target points at different distances within a measurement distance range in the image, thereby inverting the real spectral reflectivity of the object, reducing the precision of data inversion caused by uneven light source illumination and external ambient light, and improving the accuracy of spectral reflectivity inversion.

[0148] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for reconstructing the spectral reflectance of an underwater target based on differential calculation, characterized in that: The following steps are involved: S1. Perform radiation calibration on the polarization spectroscopic imager to obtain the radiation calibration coefficient, and use the Zhang Zhengyou calibration method to calculate the internal parameter matrix and distortion coefficient of the polarization spectroscopic imager. At the same time, use a standard spectral radiometer to measure the irradiance at 1 meter away from the light source; S2, placing the polarization spectrum imager and the light source at equal height underwater, turning on the light source to illuminate the target scene, and using a water body attenuation coefficient measuring instrument to measure the current water body attenuation coefficient; S3, using a polarization spectral imager to photograph the target scene when the light source is turned on and off, respectively, to obtain two sets of multispectral image sequences; S4, horizontally move the polarization spectrum imager and the light source as a whole for a certain distance, repeat S3, and obtain two sets of multispectral image sequences; S5, converting the two sets of multispectral image sequences of S3 and the two sets of multispectral image sequences of S4 into radiance values ​​before and after the movement respectively through the radiation calibration coefficient, and calculating the radiance value of the target information light of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source, and obtaining the spectral images after removing the backscattered light before and after the overall movement respectively; S6. Select spectral images of specific bands from the spectral images after removing backscattered light before and after the overall movement according to the turbidity of the water body, pre-process the two spectral images in combination with the internal parameter matrix and distortion coefficient of the polarization spectral imager, and determine the overlapping area of ​​the two images, and calculate the distance of each pixel point in the overlapping area corresponding to the target object scene from the optical center of the polarization spectral imager and the light source before the movement and the incident angle of the light source in combination with binocular ranging; S7. Reconstruct the spectral reflectance of the target scene based on the current water attenuation coefficient, the irradiance at 1 meter from the light source, the target information light radiance value before the overall movement, the distance between the target point in the target scene corresponding to each pixel in the overlapping area and the optical center of the polarization spectrum imager and the light source, and the incident angle of the light source, and complete the reconstruction of the underwater target spectral reflectance.

2. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 1 is characterized in that: In S1, the polarization mode of the polarization spectrum imager is linear polarization, and the polarization angles are 0°, 45°, 90° and 135°.

3. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 1 is characterized in that: The specific process of S5 is as follows: S5.1, take S3 two groups of multispectral image sequences as: turn on the light source, the multispectral image sequence is Turn off the light source and the multispectral image sequence is S4 Two sets of multispectral image sequences are represented as follows: Turn on the light source, and the multispectral image sequence is Turn off the light source and the multispectral image sequence is Among them, x, y represent the horizontal and vertical coordinates of the spectral image sequence, λ represents the incident wavelength, represents the polarization angle; S5.2, the four multispectral image sequences in S5.1 are converted into radiance values ​​by using the radiation calibration coefficients, which are expressed as: and S5.3, respectively subtracting the multi-spectral image sequences of the light source turned on and turned off at the same position to obtain the spectral image sequence of the active light source at the position; The backscattered light of the active light source spectral image sequence is removed to obtain a new spectral image. The target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectral imager and the light source is calculated respectively, that is, the spectral image after removing the backscattered light before and after the overall movement is obtained.

4. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 3 is characterized in that: In S5.2, the formula for converting the four multispectral image sequences in S5.1 into radiance values ​​by using the radiation calibration coefficient is: Among them, x is one of camera-light-L, camera-L, camera-light-R, camera-R; C(λ) represents the radiation calibration coefficient.

5. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 4 is characterized in that: In S5.3, the multi-spectral image sequences of turning on and off the light source at the same position are subtracted to obtain the spectral image sequence of the active light source at the position: The multi-spectral image sequences of the polarization spectral imager and the light source at the position before the overall movement of the light source are subtracted to obtain the active light source spectral image sequence L at that position. light-L (x,y,λ) is: L light-L (x,y,λ)=L camera-light-L (x,y,λ)-L camera-L (x,y,λ) After the polarization spectral imager and the light source are moved as a whole, the multi-spectral image sequences of the light source turned on and off at the position are subtracted to obtain the active light source spectral image sequence L at that position. light-R (x,y,λ) is: L light-R (x,y,λ)=L camera-light-R (x,y,λ)-L camera-R (x,y,λ).

6. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 5 is characterized in that: In S5.3, the target information light radiance values ​​of the target scene entering the imaging spectrometer before and after the polarization spectrum imager and the light source move as a whole are calculated as follows: The Stokes vector of the target information light radiance value of the target scene entering the imaging spectrometer before and after the overall movement of the polarization spectrum imager and the light source is calculated as follows: Among them, S m In (x, y, λ), m is light-L and light-R, which respectively represent the target information light radiance value of the target scene entering the imaging spectrometer before the polarization spectrum imager and the light source move as a whole, and the target information light radiance value of the target scene entering the imaging spectrometer after the polarization spectrum imager and the light source move as a whole; p scat (λ) represents the polarization degree of the backscattered light of the active light source spectrum image sequence, represents the optimal target information light polarization degree p in the active light source spectrum image sequence obj (λ); I m (x,y,λ)=L m (x,y,λ,90°)+L m (x,y,λ,0°); Q m (x,y,λ)=L m (x,y,λ,90°)-L m (x,y,λ,0°)? U m (x,y,λ)=L m (x,y,λ,45°)-L m (x,y,λ,135°); in, Represents the active light source spectrum image sequence L m (x,y,λ) The image obtained by polarization of the angle; The calculation process of the polarization degree of backscattered light of the active light source spectrum image sequence is: In the active light source spectrum image sequence corresponding to each wavelength, a target-free area σ is framed. void (λ), the degree of polarization of backscattered light of pixels in the active light source spectrum image sequence is estimated as: B max (x,y,λ) is σ void The maximum backscattered light radiance in (λ), B min (x,y,λ) is σ void (λ) is the minimum backscattered light radiance; n is σ void The number of pixels in (λ); The polarization degree p of the target information light in the active light source spectrum image sequence is obj (λ) takes values ​​from 0 to 1 in steps of 0.01, and frames the region containing a clear single target in the active light source spectrum image sequence at each wavelength. clear (λ), the optimal polarization value of the target information light is calculated using the mutual information of the target information light radiance value and the backscattered light radiance value. The formula is: In the formula, MI[B,S] represents B m (x,y,λ) and S m The mutual information value of (x, y, λ), B m (x, y, λ) represents the backscattered light radiance value, represents the optimal p obj (λ).

7. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 3 is characterized in that: S6 is specifically: S6.

1. Under a certain water turbidity, select two multispectral image sequences with the highest contrast in the spectrum before and after the overall movement after removing the backscattered light, and use the spectrum image before the movement as the left image; and use the spectrum image after the movement as the right image; S6.2, using the intrinsic parameter matrix and distortion coefficient of the polarization spectral imager and the pose information of the polarization spectral imager when shooting the target scene, perform epipolar correction on the left image and the right image after removing the backscattered light, perform stereo matching of the two images by using the SIFT operator, and determine the overlapping area of ​​the two spectral images; S6.

3. Calculate the depth Z of the target point P in the target scene corresponding to each pixel in the overlapping area p ; S6.4, perform binocular visual ranging on the left image and right image after removing the backscattered light, and combine the depth Z p , the three-dimensional geometric model calculates the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the optical center of the polarization spectrum imager before the overall movement camera (x, y), the distance l between the target point P in the target scene corresponding to each pixel point in the overlapping area and the light source before the overall movement light (x,y) and the incident angle of the light source.

8. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 7 is characterized in that: S6.3 specifically states: The optical center of the polarization spectrum imager before moving is taken as the origin O Cl , take the optical center of the mobile polarization spectrum imager as the origin O Cr , establish two coordinate systems with the moving direction as the x-axis, the vertical upward direction as the z-axis, and any direction perpendicular to the x-axis and z-axis as the y-axis; The horizontal moving distance l of the polarization spectrum imager is known L-R The horizontal coordinate values ​​of the target point P in the target scene in the overlapping area of ​​the left image and the right image are x l 、x r , the focal length f of the polarization spectrum imager, calculate the depth Z of the target point P along the z-axis P for:

9. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 8, characterized in that: In S6.4, the incident angle of the light source is the angle between the line between the light source and the target point P in the target scene before the overall movement and the angle between the target point P in the target scene at x0 Cl The angle formed by the line connecting the projections on the y plane.

10. The underwater target spectral reflectance reconstruction method based on differential calculation according to claim 7, characterized in that: The specific process of S7 is as follows: S7.1, according to the Lambert-Beer law: Among them, S light-L (x, y, λ) represents the target information light radiance value before the overall movement, L reflect (x, y, λ) represents the radiance reflected by the target in the overlapping area under the active light before the overall movement, α(λ) represents the current water attenuation coefficient, l camera (x, y) represents the distance between the target point on the target object and the optical center of the polarization spectrum imager corresponding to each pixel in the overlapping area; S7.

2. When the light source is irradiated to each point on the spherical surface at a distance of 1m, the irradiance received by each point is the same and the irradiance received by the distant target is related to the square of the irradiation distance. Combining the Lambert cosine theorem and the Lambert-Beer law, the irradiance E of the target scene in the overlapping area irradiated by the light source can be obtained. incidence (x,y,λ) is: Where E0(λ) represents the irradiance at 1 meter from the light source, θ i (x,y) represents the incident angle of the light source; l light (x, y) represents the distance between the target point and the light source on the target object corresponding to each pixel in the overlapping area; S7.

3. Calculate the spectral reflectance of the reconstructed target scene using the formula:

Images