CN109141638B - A kind of underwater polarization imaging method of natural light - Google Patents
A kind of underwater polarization imaging method of natural light Download PDFInfo
- Publication number
- CN109141638B CN109141638B CN201810827464.8A CN201810827464A CN109141638B CN 109141638 B CN109141638 B CN 109141638B CN 201810827464 A CN201810827464 A CN 201810827464A CN 109141638 B CN109141638 B CN 109141638B
- Authority
- CN
- China
- Prior art keywords
- image
- luminous intensity
- channel image
- polarization
- light
- 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
- 230000010287 polarization Effects 0.000 title claims abstract description 284
- 238000003384 imaging method Methods 0.000 title claims abstract description 36
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 description 19
- 230000003287 optical effect Effects 0.000 description 19
- 238000002310 reflectometry Methods 0.000 description 13
- 238000000034 method Methods 0.000 description 8
- 238000010521 absorption reaction Methods 0.000 description 7
- 230000003595 spectral effect Effects 0.000 description 6
- 241001062009 Indigofera Species 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 239000002245 particle Substances 0.000 description 3
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 2
- 241000282326 Felis catus Species 0.000 description 2
- 230000031700 light absorption Effects 0.000 description 2
- 210000001525 retina Anatomy 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- BTCSSZJGUNDROE-UHFFFAOYSA-N gamma-aminobutyric acid Chemical compound NCCCC(O)=O BTCSSZJGUNDROE-UHFFFAOYSA-N 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 230000000149 penetrating effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/447—Polarisation spectrometry
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Color Television Image Signal Generators (AREA)
Abstract
The present invention relates to a kind of underwater polarization imaging methods of natural light, comprising: the first model of third polarization image is established by the first polarization image and third polarization image；The second model of the 4th polarization image is established by the second polarization image and the 4th polarization image；The luminous intensity of the 5th polarization image is obtained according to the first model of the luminous intensity of third polarization image and third polarization image；The luminous intensity of the 6th polarization image is obtained according to the second model of the luminous intensity of the 4th polarization image and the 4th polarization image；The luminous intensity of final underwater picture target emanation light is obtained according to the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image.The reason of present invention is by research underwater picture cross-color, obtain the luminous intensity of the 5th polarization image and the luminous intensity of the 6th polarization image, to the color information of restoration scenario, later using the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image, the clearly final underwater picture of achromatization distortion is rebuild.
Description
Technical field
The invention belongs to optical image technology fields, and in particular to a kind of underwater polarization imaging method of natural light.
Background technique
Passive underwater polarization imaging is mainly used for the neritic province domain that sunlight can be irradiated to, and passes through the polarization using scene
Information can effectively remove backscatter light, rebuild clear scene image.Under water in scene, the factor master of image quality is influenced
Will for suspended particles in water to the scattering of light wave and water body to the absorption of light wave.Wherein the scattering of light wave can make incident light
Line is propagated towards detector direction, is superimposed upon on target information light, picture contrast is caused to reduce；The absorption of light wave can make difference
The attenuation of the light wave of wavelength in water is different, is distorted the image color information obtained.
Underwater passive optical imaging method mainly has a dark imaging method at present, Retinex (Retinex be by
The compound word that retina (retina)+cortex (cortex) is constituted, indicates a kind of computational theory of color constancy consciousness) imaging
Method and underwater polarization imaging method etc., wherein underwater polarization imaging method is due to having, device structure is simple, imaging effect
The features such as good and cost performance is high is widely used.Underwater polarization imaging method is polarized under water by obtaining two orthogonal width of polarization state
Image is isolated using backscatter light and the difference of target information light, effectively obtains clear scene image.
But although underwater polarization imaging method can remove the influence of backscatter light, imitated since water body absorbs
Cross-color problem caused by answering is not resolved, limited so as to cause the reconstruction effect of obtained underwater picture, is easy
Cause identification error.
Summary of the invention
In order to solve the above-mentioned problems in the prior art, the present invention provides a kind of natural lights to polarize image space under water
Method.
An embodiment provides a kind of underwater polarization imaging methods of natural light, comprising:
The first model of the third polarization image is established by the first polarization image and third polarization image；
The second model of the 4th polarization image is established by the second polarization image and the 4th polarization image；
The 5th polarization is obtained according to the first model of the luminous intensity of the third polarization image and the third polarization image
The luminous intensity of image；
The 6th polarization is obtained according to the second model of the luminous intensity of the 4th polarization image and the 4th polarization image
The luminous intensity of image；
Final underwater figure is obtained according to the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image
As the luminous intensity of target emanation light.
In one embodiment of the invention, which is characterized in that first polarization image include the first red channel image,
First green channel image and the first blue channel image, second polarization image include the second red channel image, the second green channel
Image and the second blue channel image, the third polarization image include the red channel image of third, the green channel image of third and third
Blue channel image, the 4th polarization image include the 4th red channel image, the 4th green channel image and the 4th blue channel image.
In one embodiment of the invention, to establish the third by the first polarization image and third polarization image inclined
Before first model of vibration image, further includes:
Described in reflection constant acquisition by the luminous intensity and the first red channel image of the described first red channel image
The light source response rate of first red channel image；
Described in reflection constant acquisition by the luminous intensity and the first green channel image of the described first green channel image
The light source response rate of first green channel image；
By described in the luminous intensity of the first blue channel image and the reflection constant acquisition of the first blue channel image
The light source response rate of first blue channel image.
In one embodiment of the invention, the third is established by the first polarization image and third polarization image to polarize
First model of image, comprising:
Pass through the reflection constant of the described first red channel image, the light source response rate of the first red channel image and described
The luminous intensity of the red channel image of third establishes the first model of the red channel image of the third；
Pass through the reflection constant of the described first green channel image, the light source response rate of the first green channel image and described
The luminous intensity of the green channel image of third establishes the first model of the green channel image of the third；
Pass through the reflection constant of the first blue channel image, the light source response rate of the first blue channel image and described
The luminous intensity of third blue channel image establishes the first model of the third blue channel image.
In one embodiment of the invention, the described 4th is being established partially by the second polarization image and the 4th polarization image
Before second model of vibration image, further includes:
Described in reflection constant acquisition by the luminous intensity and the second red channel image of the described second red channel image
The light source response rate of second red channel image；
Described in reflection constant acquisition by the luminous intensity and the second green channel image of the described second green channel image
The light source response rate of second green channel image；
By described in the luminous intensity of the second blue channel image and the reflection constant acquisition of the second blue channel image
The light source response rate of second blue channel image.
In one embodiment of the invention, the 4th polarization is established by the second polarization image and the 4th polarization image
Second model of image, comprising:
Pass through the reflection constant of the described second red channel image, the light source response rate of the second red channel image and described
The luminous intensity of 4th red channel image establishes the second model of the 4th red channel image；
Pass through the reflection constant of the described second green channel image, the light source response rate of the second green channel image and described
The luminous intensity of 4th green channel image establishes the second model of the 4th green channel image；
Pass through the reflection constant of the second blue channel image, the light source response rate of the second blue channel image and described
The luminous intensity of 4th blue channel image establishes the second model of the 4th blue channel image.
In one embodiment of the invention, according to the luminous intensity of the third polarization image and the third polarization image
The first model obtain the luminous intensity of the 5th polarization image, comprising:
The 5th is obtained according to the first model of the luminous intensity of the red channel image of the third and the red channel image of the third
The luminous intensity of red channel image；
The 5th is obtained according to the first model of the luminous intensity of the green channel image of the third and the green channel image of the third
The luminous intensity of green channel image；
The 5th is obtained according to the first model of the luminous intensity of the third blue channel image and the third blue channel image
The luminous intensity of blue channel image；
Pass through the luminous intensity of the 5th red channel image, the luminous intensity of the 5th green channel image and the 5th indigo plant
The luminous intensity of channel image obtains the luminous intensity of the 5th polarization image.
In one embodiment of the invention, according to the luminous intensity of the 4th polarization image and the 4th polarization image
The second model obtain the luminous intensity of the 6th polarization image, comprising:
The 6th is obtained according to the first model of the luminous intensity of the 4th red channel image and the 4th red channel image
The luminous intensity of red channel image；
The 6th is obtained according to the first model of the luminous intensity of the 4th green channel image and the 4th green channel image
The luminous intensity of green channel image；
The 6th is obtained according to the first model of the luminous intensity of the 4th blue channel image and the 4th blue channel image
The luminous intensity of blue channel image；
Pass through the luminous intensity of the 5th red channel image, the luminous intensity of the 5th green channel image and the 5th indigo plant
The luminous intensity of channel image obtains the luminous intensity of the 5th polarization image.
In one embodiment of the invention, according to the luminous intensity of the 5th polarization image and the 6th polarization image
Luminous intensity obtain the luminous intensity of final underwater picture target emanation light, comprising:
The polarization of the total light intensity degree of the final underwater picture, the backscatter light of the final underwater picture is obtained respectively
It spends, infinite point backscatter light in the luminous intensity of the backscatter light of the final underwater picture, the final underwater picture
Luminous intensity；
Pass through the luminous intensity of the 5th polarization image, the luminous intensity of the 6th polarization image, the final underwater figure
The total light intensity degree of picture, the final underwater picture backscatter light degree of polarization, the final underwater picture backscatter
The luminous intensity of infinite point backscatter light establishes the final underwater picture mesh in the luminous intensity of light, the final underwater picture
Mark the computation model of the luminous intensity of radiant light；
Final underwater picture target is obtained according to the computation model of the luminous intensity of the final underwater picture target emanation light
The luminous intensity of radiant light.
In one embodiment of the invention, the computation model of the luminous intensity of the final underwater picture target emanation light
Are as follows:
Wherein, IObjectIt (x) is the luminous intensity of final underwater picture target emanation light, IToatlIt (x) is final underwater picture
Total light intensity degree, I'maxIt (x) is the luminous intensity of the 5th polarization image, I'minIt (x) is the luminous intensity of the 6th polarization image, p is final
The degree of polarization of the backscatter light of underwater picture, IBIt (x) is the luminous intensity of the backscatter light of final underwater picture,For
The luminous intensity of infinite point backscatter light in final underwater picture.
Compared with prior art, beneficial effects of the present invention:
The reason of present invention is by research underwater picture cross-color, obtains the 5th polarization image based on depth information
The luminous intensity of luminous intensity and the 6th polarization image, so that the color information of restoration scenario, utilizes the light of the 5th polarization image later
The luminous intensity of intensity and the 6th polarization image rebuilds the clearly final underwater picture of achromatization distortion.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the underwater polarization imaging method of natural light provided in an embodiment of the present invention.
Specific embodiment
Further detailed description is done to the present invention combined with specific embodiments below, but embodiments of the present invention are not limited to
This.
Embodiment one
Referring to Figure 1, Fig. 1 is a kind of process signal of underwater polarization imaging method of natural light provided in an embodiment of the present invention
Figure.A kind of underwater polarization imaging method of natural light provided in an embodiment of the present invention, comprising:
The first model of the third polarization image is established by the first polarization image and third polarization image；
The second model of the 4th polarization image is established by the second polarization image and the 4th polarization image；
The 5th polarization is obtained according to the first model of the luminous intensity of the third polarization image and the third polarization image
The luminous intensity of image；
The 6th polarization is obtained according to the second model of the luminous intensity of the 4th polarization image and the 4th polarization image
The luminous intensity of image；
Final underwater figure is obtained according to the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image
As the luminous intensity of target emanation light.
The reason of the present embodiment is by research underwater picture cross-color, obtains the 5th polarization image based on depth information
Luminous intensity and the luminous intensity of the 6th polarization image utilize the 5th polarization image later thus the color information of restoration scenario
The luminous intensity of luminous intensity and the 6th polarization image rebuilds the clearly final underwater picture of achromatization distortion.
Embodiment two
The embodiment of the present invention is on the basis of the above embodiments, inclined under water to a kind of natural light provided in an embodiment of the present invention
Vibration imaging method is specifically introduced, and this method specifically includes:
Step 1: establishing the first model of third polarization image by the first polarization image and third polarization image；
Step 1.1, the luminous intensity for obtaining the first polarization image and the second polarization image respectively；
Blue keynote is generally all presented in underwater picture, and the underwater picture that blue keynote is presented can be regarded as in blue light
The image of lower shooting, therefore estimate that light source colour can remove blue keynote.
In an atmosphere, detector is made to obtain two orthogonal width polarization images of polarization state, two width polarization by rotatory polarization piece
Image includes the first polarization image and the second polarization image, and the luminous intensity of the first polarization image of note isSecond polarization image
Luminous intensity beWherein, the luminous intensity of backscatter light possessed by the first polarization image is maximum, and just with its polarization state
The second polarization image handed over has the luminous intensity of the smallest backscatter light.
Specifically, the first polarization image is divided into the first red channel image, the first green channel figure according to red, green and blue triple channel
Second polarization image is divided into the second red channel image, second green according to red, green and blue triple channel by picture and the first blue channel image
Channel image and the second blue channel image.
The luminous intensity for obtaining the first polarization image obtains luminous intensity, the first green channel figure of the first red channel image respectively
The luminous intensity of the luminous intensity of picture and the first blue channel image, it is red logical that the luminous intensity of the second polarization image of acquisition obtains second respectively
The luminous intensity of the luminous intensity of road image, the luminous intensity of the second green channel image and the second blue channel image.
Step 1.1.1, the luminous intensity of the first red channel image is obtained according to Lambertian (lambert) reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point (namely
The color of certain point in the picture that camera obtains), shown in the luminous intensity such as formula (1) of the first red channel image:
Wherein,For the luminous intensity of the first red channel image, λ1For the optical wavelength of the first red channel image, ω1
For the spectral region of the first red channel image, x is the position of pixel x in the first red channel image, e1(λ1) it is first red logical
The distribution of light source, s in road image1(x,λ1) it is pixel x in the first red channel image to optical wavelength λ1Reflectivity, g1(λ1)
For the corresponding photosensitive coefficient of camera of the first red channel image.
Step 1.1.2, the luminous intensity of the first green channel image is obtained according to Lambertian reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point, first is green
Shown in the luminous intensity of channel image such as formula (2):
Wherein,For the luminous intensity of the first green channel image, λ2For the optical wavelength of the first green channel image, ω2
For the spectral region of the first green channel image, x is the position of pixel x in the first green channel image, e2(λ2) it is first green logical
The distribution of light source, s in road image2(x,λ2) it is pixel x in the first green channel image to optical wavelength λ2Reflectivity, g2(λ2)
For the corresponding photosensitive coefficient of camera of the first green channel image.
Step 1.1.3, the luminous intensity of the first blue channel image is obtained according to Lambertian reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point, first is blue
Shown in the luminous intensity of channel image such as formula (3):
Wherein,For the luminous intensity of the first blue channel image, λ3For the optical wavelength of the first blue channel image, ω3
For the spectral region of the first blue channel image, x is the position of pixel x in the first blue channel image, e3(λ3) it is first blue logical
The distribution of light source, s in road image3(x,λ3) it is pixel x in the first blue channel image to optical wavelength λ3Reflectivity, g3(λ3)
For the corresponding photosensitive coefficient of camera of the first blue channel image.
Step 1.1.4, the luminous intensity of the second red channel image is obtained according to Lambertian reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point, second is red
Shown in the luminous intensity of channel image such as formula (4):
Wherein,For the luminous intensity of the second red channel image, λ4For the optical wavelength of the second red channel image, ω4
For the spectral region of the second red channel image, x is the position of pixel x in the second red channel image, e4(λ4) it is second red logical
The distribution of light source, s in road image4(x,λ4) it is pixel x in the second red channel image to optical wavelength λ4Reflectivity, g4(λ4)
For the corresponding photosensitive coefficient of camera of the second red channel image.
Step 1.1.5, the luminous intensity of the second green channel image is obtained according to Lambertian reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point, second is green
Shown in the luminous intensity of channel image such as formula (5):
Wherein,For the luminous intensity of the second green channel image, λ5For the optical wavelength of the second green channel image, ω5
For the spectral region of the second green channel image, x is the position of pixel x in the second green channel image, e5(λ5) it is second green logical
The distribution of light source, s in road image5(x,λ5) it is pixel x in the second green channel image to optical wavelength λ5Reflectivity, g5(λ5)
For the corresponding photosensitive coefficient of camera of the second green channel image.
Step 1.1.2, the luminous intensity of the second blue channel image is obtained according to Lambertian reflection model；
By Lambertian reflection model it is found that in atmospheric sounding scene on body surface when the color of certain point, second is blue
Shown in the initial light intensity of channel image such as formula (6):
Wherein,For the luminous intensity of the second blue channel image, λ6For the optical wavelength of the second blue channel image, ω6
For the spectral region of the second blue channel image, x is the position of pixel x in the second blue channel image, e6(λ6) it is second blue logical
The distribution of light source, s in road image6(x,λ6) it is pixel x in the second blue channel image to optical wavelength λ6Reflectivity, g6(λ6)
For the corresponding photosensitive coefficient of camera of the second blue channel image.
Step 1.2, the reflection constant for obtaining the first polarization image and the second polarization image respectively；
It obtains the reflection constant of the first polarization image and obtains the reflection constant of the first red channel image, first green logical respectively
The reflection constant of the reflection constant of road image and the first blue channel image, the reflection constant for obtaining the second polarization image obtain respectively
Take the reflection of the reflection constant of the second red channel image, the reflection constant of the second green channel image and the second blue channel image normal
Number.
Step 1.2.1, the reflection constant of the first red channel image is obtained；
From physical significance, Gray Word (grey world) assumes nature scenery for the average reflection of light
Mean value be definite value on the whole, this definite value is approximately " grey ", and Gray Word hypothesis thinks: property in scene
The average reflection in body surface face is no color differnece, i.e., according to the available reflection constant of reflectivity, the reflection of the first red channel image
Shown in constant such as formula (7):
∫s1(x,λ1) dx/ ∫ dx=k1 (7)
Wherein, k1 is the reflection constant of the first red channel image, k1A constant between [0,1], s1(x,λ1) be
Pixel x is to optical wavelength λ in first red channel image1Reflectivity.
Step 1.2.2, the reflection constant of the first green channel image is obtained；
Shown in the reflection constant such as formula (8) of first green channel image:
∫s2(x,λ2) dx/ ∫ dx=k2 (8)
Wherein, k2For the reflection constant of the first green channel image, k2A constant between [0,1], s2(x,λ2) be
Pixel x is to optical wavelength λ in first green channel image2Reflectivity.
Step 1.2.3, the reflection constant of the first blue channel image is obtained；
Shown in the reflection constant such as formula (9) of first blue channel image:
∫s3(x,λ3) dx/ ∫ dx=k3 (9)
Wherein, k3For the reflection constant of the first blue channel image, k3A constant between [0,1], s3(x,λ3) be
Pixel x is to optical wavelength λ in first blue channel image3Reflectivity.
Step 1.2.4, the reflection constant of the second red channel image is obtained；
Shown in the reflection constant such as formula (10) of second red channel image:
∫s4(x,λ4) dx/ ∫ dx=k4 (10)
Wherein, k4For the reflection constant of the second red channel image, k4A constant between [0,1], s4(x,λ4) be
Pixel x is to optical wavelength λ in second red channel image4Reflectivity.
Step 1.2.5, the reflection constant of the second green channel image is obtained；
Shown in the reflection constant such as formula (11) of second green channel image:
∫s5(x,λ5) dx/ ∫ dx=k5 (11)
Wherein, k5For the reflection constant of the second green channel image, k5A constant between [0,1], s5(x,λ5) be
Pixel x is to optical wavelength λ in second green channel image5Reflectivity.
Step 1.2.6, the reflection constant of the second blue channel image is obtained；
Shown in the reflection constant such as formula (12) of second blue channel image:
∫s6(x,λ6) dx/ ∫ dx=k6 (12)
Wherein, k6For the reflection constant of the second blue channel image, k6A constant between [0,1], s6(x,λ6) be
Pixel x is to optical wavelength λ in second blue channel image6Reflectivity.
Step 1.3, the light source response rate for obtaining the first polarization image and the second polarization image respectively；
The light source response rate for obtaining the first polarization image obtains the light source response rate of the first red channel image, first respectively
The light source response rate of the light source response rate of green channel image and the first blue channel image obtains the light source response of the second polarization image
Rate obtains light source response rate, the light source response rate and the second blue channel of the second green channel image of the second red channel image respectively
The light source response rate of image.
Step 1.3.1, the light source response rate of the first red channel image is obtained；
Specifically, red by the acquisition first of the reflection constant of the luminous intensity of the first red channel image and the first red channel image
The light source response rate of channel image, according to formula (1) and formula (7) it can be concluded that the luminous intensity and first of the first red channel image
Shown in the relationship of the reflection constant of red channel image such as formula (13):
Wherein, E1=e1(λ1)g1(λ1), E1For the light source response rate of the first red channel image.
Step 1.3.2, the light source response rate of the first green channel image is obtained；
Specifically, green by the acquisition first of the reflection constant of the luminous intensity of the first green channel image and the first green channel image
The light source response rate of channel image.According to formula (2) and formula (8) it can be concluded that the luminous intensity and first of the first green channel image
The relationship of the reflection constant of green channel image, as shown in formula (14):
Wherein, E2=e2(λ2)g2(λ2), E2 is the light source response rate of the first green channel image.
Step 1.3.3, the light source response rate of the first blue channel image is obtained；
Specifically, green by the luminous intensity of the first blue channel image and the acquisition first of the reflection constant of the first blue channel image
The light source response rate of channel image.According to formula (3) and formula (9) it can be concluded that the luminous intensity and first of the first blue channel image
The relationship of the reflection constant of blue channel image, as shown in formula (15):
Wherein, E3=e3(λ3)g3(λ3), E3For the light source response rate of the first blue channel image.
Step 1.3.4, the light source response rate of the second red channel image is obtained；
The second red channel figure is obtained by the luminous intensity of the second red channel image and the reflection constant of the second red channel image
The light source response rate of picture.According to formula (4) and formula (10) it can be concluded that in the second red channel image pixel x luminous intensity
With the relationship of the reflection constant of the second red channel image, as shown in formula (16):
Wherein, E4=e4(λ4)g4(λ4), E4For the light source response rate of the second red channel image.
Step 1.3.5, the light source response rate of the second green channel image is obtained；
The second green channel figure is obtained by the luminous intensity of the second green channel image and the reflection constant of the second green channel image
The light source response rate of picture.According to formula (5) and formula (11) it can be concluded that in the second green channel image pixel x luminous intensity
With the relationship of the reflection constant of the second green channel image, as shown in formula (17):
Wherein, E5=e5(λ5)g5(λ5), E5For the light source response rate of the second green channel image.
Step 1.3.6, the light source response rate of the second blue channel image is obtained；
The second green channel figure is obtained by the luminous intensity of the second blue channel image and the reflection constant of the second blue channel image
The light source response rate of picture.According to formula (6) and formula (12) it can be concluded that in the second blue channel image pixel x luminous intensity
With the relationship of the reflection constant of the second blue channel image, as shown in formula (18):
Wherein, E6=e6(λ6)g6(λ6), E6For the light source response rate of the second blue channel image.
WithIt is to shoot institute in an atmosphere
The luminous intensity of the image of acquisition, above-mentioned algorithm is causing the image processing effect of cross-color good because of light source, but ring under water
The image attenuation degree that border obtains is serious more than in an atmosphere, and above-mentioned algorithm correction of color can not be distorted well.
Step 1.4, the luminous intensity for obtaining third polarization image and the 4th polarization image respectively；
Under water in the imaging process of polarization imaging technology, detector is set to obtain polarization state by rotatory polarization piece orthogonal
Two width polarization images, two width polarization images include third polarization image and the 4th polarization image, the luminous intensity of third polarization image
ForThe initial light intensity of 4th polarization image isWherein, the light of backscatter light possessed by third polarization image
Maximum intensity, and fourth polarization image orthogonal with its polarization state has the luminous intensity of the smallest backscatter light.Third polarization
Image and the 4th polarization image are the underwater picture of cross-color.
By third polarization image according to red, green and blue triple channel be divided into the red channel image of third, the green channel image of third and
4th polarization image is divided into the 4th red channel image, the 4th green channel according to red, green and blue triple channel by third blue channel image
Image and the 4th blue channel image.
The luminous intensity for obtaining third polarization image obtains the green channel figure of luminous intensity, third of the red channel image of third respectively
The luminous intensity of picture and the luminous intensity of third blue channel image, it is red logical that the luminous intensity of the 4th polarization image of acquisition obtains the 4th respectively
The luminous intensity of the luminous intensity of road image, the luminous intensity of the 4th green channel image and the 4th blue channel image.
Step 1.4.1, the luminous intensity of the red channel image of third is obtained；
Shown in the luminous intensity such as formula (19) of the red channel image of the third of cross-color:
Wherein,For the luminous intensity of the red channel image of third,For the light wave of the red channel image of third
Decay factor,I.e.For the decay factor of the light wave of the red channel image of third,It is red for third
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f1It (x) is red logical by obtained in air first
The luminous intensity of road image,
Step 1.4.2, the luminous intensity of the green channel image of third is obtained；
Shown in the luminous intensity such as formula (20) of the green channel image of the third of cross-color:
Wherein,For the luminous intensity of the green channel image of third,For the light wave of the green channel image of third
Decay factor,I.e.For the decay factor of the light wave of the green channel image of third,It is green for third
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f2It (x) is green logical by obtained in air first
The luminous intensity of road image,
Step 1.4.3, the luminous intensity of third blue channel image is obtained；
Shown in the luminous intensity such as formula (21) of the third blue channel image of cross-color:
Wherein,For the luminous intensity of third blue channel image,For the light wave of third blue channel image
Decay factor,I.e.For the decay factor of the light wave of third blue channel image,For third indigo plant
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f3It (x) is blue logical by obtained in air first
The luminous intensity of road image,
Step 1.4.4, the luminous intensity of the 4th red channel image is obtained；
Shown in the luminous intensity such as formula (22) of 4th red channel image of cross-color:
Wherein,For the luminous intensity of the 4th red channel image,For the light wave of the 4th red channel image
Decay factor,I.e.For the decay factor of the light wave of the 4th red channel image,It is red for the 4th
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f4It (x) is red logical by obtained in air second
The luminous intensity of road image,
Step 1.4.5, the luminous intensity of the 4th green channel image is obtained；
Shown in the luminous intensity such as formula (23) of 4th green channel image of cross-color:
Wherein,For the luminous intensity of the 4th green channel image,For the light wave of the 4th green channel image
Decay factor,I.e.For the decay factor of the light wave of the 4th green channel image,It is green for the 4th
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f5It (x) is green logical by obtained in air second
The luminous intensity of road image,
Step 1.4.6, the luminous intensity of the 4th blue channel image is obtained；
Shown in the luminous intensity such as formula (24) of 4th blue channel image of cross-color:
Wherein,For the luminous intensity of the 4th blue channel image,For the light wave of the 4th blue channel image
Decay factor,I.e.For the decay factor of the light wave of the 4th blue channel image,For the 4th indigo plant
The attenuation coefficient of the light wave of channel image, z are distance of the target to detector, f6It (x) is blue logical by obtained in air second
The luminous intensity of road image,
The decay factor of step 1.5, the respectively light wave of acquisition third polarization image and the 4th polarization image；
Obtain the decay factor i.e. decaying of the light wave of the acquisition red channel image of third respectively of the light wave of third polarization image
The decay factor of the light wave of the decay factor and third blue channel image of the light wave of the green channel image of the factor, third obtains the 4th
The decay factor of the light wave of polarization image obtains the decay factor of the light wave of the 4th red channel image, the 4th green channel figure respectively
The decay factor of the light wave of the decay factor of the light wave of picture and the 4th blue channel image.
Step 1.5.1, the decay factor of the light wave of third blue channel image is obtained；
In the red channel image of third the scattering coefficient of the luminous intensity of infinite point backscatter and the red channel image of third at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of the red channel image of third, as shown in formula (25):
Wherein,For the luminous intensity of infinite point backscatter in the red channel image of third,For the red channel figure of third
The scattering coefficient of picture,For the attenuation coefficient of the light wave of the red channel image of third.
In the green channel image of third the scattering coefficient of the luminous intensity of infinite point backscatter and the green channel image of third at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of the red channel image of third, as shown in formula (26):
Wherein,For the luminous intensity of infinite point backscatter in the green channel image of third,For the green channel figure of third
The scattering coefficient of picture,For the attenuation coefficient of the light wave of the green channel image of third.
In third blue channel image the scattering coefficient of the luminous intensity of infinite point backscatter and third blue channel image at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of third blue channel image, as shown in formula (27):
Wherein,For the luminous intensity of infinite point backscatter in third blue channel image,For third blue channel figure
The scattering coefficient of picture,For the attenuation coefficient of the light wave of third blue channel image.
Under water in imaging process, the attenuation degree of blue light is minimum, therefore can be using blue light as reference, thus respectively
Find out the attenuation ratio of red light, green light relative to blue light, wherein the red channel image of third is relative to third blue channel image
Decaying such as formula (28) shown in:
Shown in decaying such as formula (29) of the green channel image of third relative to third blue channel image:
The luminous intensity of the backscatter light of third blue channel image can be obtained according to the luminous intensity calculation formula of backscatter light
As shown in formula (30):
Wherein,For the luminous intensity of the backscatter light of third blue channel image.
Assuming that the decaying of blue light is only related with the depth of field, according to formula (21) and the available third blue channel of formula (30)
Shown in the decay factor of the light wave of image such as formula (31):
Step 1.5.2, the decay factor of the light wave of the red channel image of third is obtained；
In the identical situation of propagation distance of third blue channel image and the red channel image of third, according toThe decay factor of formula (28) and the light wave of the red channel image of formula (31) available third is such as public
Shown in formula (32):
Step 1.5.3, the decay factor of the light wave of the green channel image of third is obtained；
In the identical situation of propagation distance of third blue channel image and the green channel image of third, according toThe decay factor of formula (29) and the light wave of the green channel image of formula (31) available third is such as public
Shown in formula (33):
Step 1.5.4, the decay factor of the light wave of the 4th blue channel image is obtained；
In 4th red channel image the scattering coefficient of the luminous intensity Yu the 4th red channel image of infinite point backscatter at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of the 4th red channel image, as shown in formula (34):
Wherein,For the luminous intensity of infinite point backscatter in the 4th red channel image,For the 4th red channel figure
The scattering coefficient of picture,For the attenuation coefficient of the light wave of the 4th red channel image.
In 4th green channel image the scattering coefficient of the luminous intensity Yu the 4th green channel image of infinite point backscatter at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of the 4th green channel image, as shown in formula (35):
Wherein,For the luminous intensity of infinite point backscatter in the 4th green channel image,For the 4th green channel figure
The scattering coefficient of picture,For the attenuation coefficient of the light wave of the 4th green channel image.
In 4th blue channel image the scattering coefficient of the luminous intensity of infinite point backscatter and the 4th blue channel image at
Direct ratio is inversely proportional with the attenuation coefficient of the light wave of the 4th blue channel image, as shown in formula (36):
Wherein,For the luminous intensity of infinite point backscatter in the 4th blue channel image,For the 4th blue channel figure
The scattering coefficient of picture,For the attenuation coefficient of the light wave of the 4th blue channel image.
Under water in imaging process, the attenuation degree of blue light is minimum, therefore can be using blue light as reference, thus respectively
Find out the attenuation ratio of red light, green light relative to blue light, wherein the 4th red channel image is relative to the 4th blue channel image
Decaying such as formula (37) shown in:
Shown in decaying such as formula (38) of the 4th green channel image relative to the 4th blue channel image:
The luminous intensity of the backscatter light of the 4th blue channel image can be obtained according to the luminous intensity calculation formula of backscatter light
As shown in formula (39):
Wherein,For the luminous intensity of the backscatter light of the 4th blue channel image.
Assuming that the decaying of blue light is only related with the depth of field, according to formula (24) and available 4th blue channel of formula (39)
Shown in the decay factor of the light wave of image such as formula (40):
Step 1.5.5, the decay factor of the light wave of the 4th red channel image is obtained；
In the identical situation of propagation distance of the 4th blue channel image and the 4th red channel image, according toThe decay factor of formula (37) and the light wave of the available 4th red channel image of formula (40) is such as public
Shown in formula (41):
Step 1.5.5, the decay factor of the light wave of the 4th green channel image is obtained；
In the identical situation of propagation distance of the 4th blue channel image and the 4th green channel image, according toThe decay factor of formula (38) and the light wave of the available 4th green channel image of formula (40) is such as public
Shown in formula (40):
Step 1.6, the first model for establishing third polarization image；
The first model for establishing third polarization image includes that the first model for establishing the red channel image of third respectively, third are green
First model of channel image and the first model of third blue channel image.
Step 1.6.1, the first model of the red channel image of third is established；
Specifically, pass through the reflection constant of the first red channel image, the light source response rate and third of the first red channel image
The luminous intensity of red channel image establishes the first model of the red channel image of third.I.e. according to formula (13), formula (19), formula
(31) and the decay factor of the light wave of the luminous intensity of the red channel image of formula (32) available third and the red channel image of third,
The relationship of the reflection constant of the light source response rate of first red channel image and the first red channel image, the relationship are that third is red logical
First model of road image, shown in the first model such as formula (43) of the red channel image of third:
Step 1.6.2, the first model of the green channel image of third is established；
Specifically, pass through the reflection constant of the first green channel image, the light source response rate and third of the first green channel image
The luminous intensity of green channel image establishes the first model of the green channel image of third.I.e. according to formula (14), formula (20), formula
(31) and the decay factor of the light wave of the luminous intensity of the green channel image of formula (33) available third and the green channel image of third,
The relationship of the reflection constant of the light source response rate of first green channel image and the first green channel image, the relationship are that third is green logical
First model of road image, shown in the first model such as formula (44) of the green channel image of third:
Step 1.6.3, the first model of third blue channel image is established；
Specifically, pass through the reflection constant of the first blue channel image, the light source response rate and third of the first blue channel image
The luminous intensity of blue channel image establishes the first model of third blue channel image.I.e. according to formula (15), formula (21) and formula
(31) decay factor of the light wave of the luminous intensity of available third blue channel image and third blue channel image, the first blue channel
The relationship of the light source response rate of image and the reflection constant of the first blue channel image, the relationship are the of third blue channel image
One model, shown in the first model such as formula (45) of third blue channel image:
Step 2: establishing the second model of the 4th polarization image by the second polarization image and the 4th polarization image；
The second model for establishing the 4th polarization image includes the second model for establishing the 4th red channel image respectively, the 4th green
Second model of channel image and the second model of the 4th blue channel image.
Step 2.1, the second model for establishing the 4th red channel image；
Specifically, pass through the reflection constant of the second red channel image, the light source response rate and the 4th of the second red channel image
The luminous intensity of red channel image establishes the second model of the 4th red channel image.I.e. according to formula (16), formula (22), formula
(40) and the decay factor of the light wave of the luminous intensity of the available 4th red channel image of formula (41) and the 4th red channel image,
The relationship of the reflection constant of the light source response rate of second red channel image and the second red channel image, the relationship are the 4th red logical
Second model of road image, shown in the second model such as formula (46) of the 4th red channel image:
Step 2.2, the second model for establishing the 4th green channel image；
Specifically, pass through the reflection constant of the second green channel image, the light source response rate and the 4th of the second green channel image
The luminous intensity of green channel image establishes the second model of the 4th green channel image.I.e. according to formula (17), formula (23), formula
(40) and the decay factor of the light wave of the luminous intensity of the available 4th green channel image of formula (42) and the 4th green channel image,
The relationship of the reflection constant of the light source response rate of second green channel image and the second green channel image, the relationship are the 4th green logical
Second model of road image, shown in the second model such as formula (47) of the 4th green channel image:
Step 2.3, the second model for establishing the 4th blue channel image；
Specifically, pass through the reflection constant of the second blue channel image, the light source response rate and the 4th of the second blue channel image
The luminous intensity of blue channel image establishes the second model of the 4th blue channel image.I.e. according to formula (18), formula (24) and formula
(40) decay factor of the light wave of the luminous intensity and the 4th blue channel image of available 4th blue channel image, the second blue channel
The relationship of the light source response rate of image and the reflection constant of the second blue channel image, the relationship are the of the 4th blue channel image
Two models, shown in the second model such as formula (48) of the 4th blue channel image:
Step 3: obtaining the 5th polarization figure according to the first model of the luminous intensity of third polarization image and third polarization image
The luminous intensity of picture；
Step 3.1 obtains the 5th according to the luminous intensity of the red channel image of third and the first model of the red channel image of third
The luminous intensity of red channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, third is red
Shown in the calculation formula of the luminous intensity of channel image such as formula (49):
The 5th red channel figure is obtained according to the first model of the luminous intensity of the red channel image of third and the red channel image of third
The luminous intensity of picture.I.e. according to the luminous intensity of formula (43) and the available 5th red channel image of formula (49), the 5th red channel
Shown in the calculation formula of the luminous intensity of image such as formula (50):
Wherein,For the luminous intensity of the 5th red channel image, the 5th red channel image expression is that removal color is lost
The underwater picture in the red channel after very.
Step 3.2 obtains the 5th according to the luminous intensity of the green channel image of third and the first model of the green channel image of third
The luminous intensity of green channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, third is green
Shown in the calculation formula of the luminous intensity of channel image such as formula (51):
The 5th green channel figure is obtained according to the first model of the luminous intensity of the green channel image of third and the green channel image of third
The luminous intensity of picture.I.e. according to the luminous intensity of formula (44) and the available 5th green channel image of formula (51), the 5th green channel
Shown in the calculation formula of the luminous intensity of image such as formula (52):
Wherein,For the luminous intensity of the 5th green channel image, the 5th green channel image expression is that removal color is lost
The underwater picture in the green channel after very.
Step 3.3 obtains the 5th according to the luminous intensity of third blue channel image and the first model of third blue channel image
The luminous intensity of blue channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, third indigo plant
Shown in the calculation formula of the luminous intensity of channel image such as formula (53):
The 5th blue channel figure is obtained according to the first model of the luminous intensity of third blue channel image and third blue channel image
The luminous intensity of picture.I.e. according to the luminous intensity of formula (45) and the available 5th blue channel image of formula (53), the 5th blue channel
Shown in the calculation formula of the luminous intensity of image such as formula (54):
Wherein,For the luminous intensity of the 5th blue channel image, the expression of the 5th blue channel image is that removal color is lost
The underwater picture of blue channel after very.
Step 3.4, the luminous intensity by the 5th red channel image, the luminous intensity and the 5th blue channel of the 5th green channel image
The luminous intensity of image obtains the luminous intensity of the 5th polarization image；
According to formula (50), (52) and (54) available 5th polarization image, the calculation formula of the 5th polarization image is such as
Shown in formula (55):
Wherein, I′maxIt (x) is the luminous intensity of the 5th polarization image, cat (abbreviation of Concatenate) is in matlab
A function, for constructing Multidimensional numerical, i.e., by the 5th red channel image, the 5th green channel image and the 5th blue channel image
The 5th polarization image is synthesized, the 5th polarization image indicates to be the underwater picture removed after cross-color.
Step 4: obtaining the 6th polarization figure according to the second model of the luminous intensity of the 4th polarization image and the 4th polarization image
The luminous intensity of picture；
Step 4.1 obtains the 6th according to the luminous intensity of the 4th red channel image and the first model of the 4th red channel image
The luminous intensity of red channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, the 4th is red
Shown in the calculation formula of the luminous intensity of channel image such as formula (56):
The 6th red channel figure is obtained according to the first model of the luminous intensity of the 4th red channel image and the 4th red channel image
The luminous intensity of picture.I.e. according to the luminous intensity of formula (46) and the available 6th red channel image of formula (56), the 6th red channel
Shown in the calculation formula of the luminous intensity of image such as formula (57):
Wherein,For the luminous intensity of the 6th red channel image, the 6th red channel image expression is that removal color is lost
The underwater picture in the red channel after very.
Step 4.2 obtains the 6th according to the luminous intensity of the 4th green channel image and the first model of the 4th green channel image
The luminous intensity of green channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, the 4th is green
Shown in the calculation formula of the luminous intensity of channel image such as formula (58):
The 6th green channel figure is obtained according to the first model of the luminous intensity of the 4th green channel image and the 4th green channel image
The luminous intensity of picture.I.e. according to the luminous intensity of formula (47) and the available 6th green channel image of formula (58), the 6th green channel
Shown in the calculation formula of the luminous intensity of image such as formula (59):
Wherein,For the luminous intensity of the 6th green channel image, the 6th green channel image expression is that removal color is lost
The underwater picture in the green channel after very.
Step 4.3 obtains the 6th according to the luminous intensity of the 4th blue channel image and the first model of the 4th blue channel image
The luminous intensity of blue channel image；
After the absorption in view of water body, the underwater Lambertian reflection model based on depth information, the 4th is blue
Shown in the calculation formula of the luminous intensity of channel image such as formula (60):
The 6th blue channel figure is obtained according to the first model of the luminous intensity of the 4th blue channel image and the 4th blue channel image
The luminous intensity of picture.I.e. according to the luminous intensity of formula (48) and the available 6th blue channel image of formula (60), the 6th blue channel
Shown in the calculation formula of the luminous intensity of image such as formula (61):
Wherein,For the luminous intensity of the 6th blue channel image, the expression of the 6th blue channel image is that removal color is lost
The underwater picture of blue channel after very.
Step 4.4, the luminous intensity by the 5th red channel image, the luminous intensity and the 5th blue channel of the 5th green channel image
The luminous intensity of image obtains the luminous intensity of the 5th polarization image；
According to formula (57), (59) and (61) available 6th polarization image, the calculation formula of the 6th polarization image is such as
Shown in formula (62):
Wherein, I′minIt (x) is the luminous intensity of the 6th polarization image, cat (abbreviation of Concatenate) is in matlab
A function, for constructing Multidimensional numerical, i.e., by the 6th red channel image, the 6th green channel image and the 6th blue channel image
The 6th polarization image is synthesized, the 6th polarization image indicates to be the underwater picture removed after cross-color.
5th polarization image and the 6th polarization image are orthogonal in polarization state, backscatter light possessed by the 5th polarization image
Luminous intensity it is maximum, and sixth polarization image orthogonal with its polarization state has the luminous intensity of the smallest backscatter light.
Step 5: obtaining final underwater picture according to the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image
The luminous intensity of target emanation light；
Step 5.1, obtain final underwater picture backscatter light luminous intensity；
When the neritic province domain of natural light irradiation carries out Underwater Imaging, information that detector receives includes two parts, and one
The information that part is received by a detector after the absorption of water body and scattering for target emanation light, referred to as target information light, it is another
Part is that natural light scatters through water body and suspended particles and reach the light of detector, referred to as backscatter light.
Shown in the calculation formula such as formula (63) of the luminous intensity of the target information light of final underwater picture:
IO(x)=IObject(x)exp(-βλz) (63)
Wherein, IOIt (x) is the luminous intensity of the target information light of final underwater picture, IObjectIt (x) is final underwater picture mesh
Mark the luminous intensity of radiant light, βλFor the attenuation coefficient of the light wave of final underwater picture, z is distance of the target to detector, exp (-
βλZ) decay factor of the light wave of final underwater picture is indicated.The luminous intensity of the target information light of final underwater picture is with z's
Increase its energy exponentially to decay.
Shown in the calculation formula such as formula (64) of the luminous intensity of the backscatter light of final underwater picture:
Wherein, IBIt (x) is the luminous intensity of the backscatter light of final underwater picture,For nothing in final underwater picture
The luminous intensity of poor distant place backscatter light, the target information light with final underwater picture is on the contrary, the background of final underwater picture dissipates
The luminous intensity for penetrating light increases as z increases.
Step 5.2, the total light intensity degree for obtaining final underwater picture；
Shown in the calculation formula such as formula (65) of the total light intensity degree for the final underwater picture that detector obtains:
Wherein, IToatlIt (x) is the total light intensity degree of final underwater picture.
Step 5.3, obtain final underwater picture backscatter light luminous intensity；
In passive underwater polarization imaging, scattering of the natural light through particle in water can have polarization characteristic.It is inclined by rotating
Vibration piece makes detector obtain two orthogonal width polarization images of polarization state, but utilizes detector institute in passive underwater polarization imaging
The phenomenon that final underwater picture obtained has cross-color is replaced using the 5th obtained polarization image and the 6th polarization image
Two width polarization images acquired in detector, can remove the cross-color of final underwater picture when passive underwater polarization imaging
Phenomenon rebuilds the clear underwater scene of achromatization distortion.
Since the target information light of final underwater picture is non-polarized light, when it passes through linear polarizer at any angle
When, the energy of the target information light of final underwater picture is filtered half, is wrapped in the 5th polarization image and the 6th polarization image
The target information light contained is identical, therefore the luminous intensity of the target information light of the luminous intensity of the 5th polarization image and final underwater picture
Shown in relationship such as formula (66) with the luminous intensity of the backscatter light of the 5th polarization image:
Wherein, I'maxIt (x) is the luminous intensity of the 5th polarization image, IOIt (x) is the target information light of final underwater picture
Luminous intensity,For the luminous intensity of the backscatter light of the 5th polarization image.
The luminous intensity and the 6th polarization image of the target information light of the luminous intensity and final underwater picture of 6th polarization image
Backscatter light luminous intensity relationship such as formula (67) shown in:
Wherein, I'minIt (x) is the luminous intensity of the 6th polarization image, IOIt (x) is the target information light of final underwater picture
Luminous intensity,For the luminous intensity of the backscatter light of the 6th polarization image.
The luminous intensity of the luminous intensity and the 6th polarization image of the total light intensity degree and the 5th polarization image of final underwater picture
Shown in relationship such as formula (68):
ITotal(x)=I'max(x)+I'min(x) (68)
Wherein, IToatlIt (x) is the total light intensity degree of final underwater picture, I′maxIt (x) is the luminous intensity of the 5th polarization image, I′minIt (x) is the luminous intensity of the 6th polarization image.
The luminous intensity of the backscatter light of final underwater picture and the luminous intensity of the 5th polarization image, the 6th polarization image
Shown in the relationship such as formula (69) of the degree of polarization of the backscatter light of luminous intensity and final underwater picture:
IB(x)=(I'max(x)-I'min(x))/p (69)
Wherein, IBIt (x) is the luminous intensity of the backscatter light of final underwater picture, I 'maxIt (x) is the 5th polarization image
Luminous intensity, I 'minIt (x) is the luminous intensity of the 6th polarization image, p is the degree of polarization of the backscatter light of final underwater picture.
Step 5.4, obtain final underwater picture backscatter light degree of polarization；
The luminous intensity of the backscatter light of the degree of polarization and the 5th polarization image of the backscatter light of final underwater picture and
Shown in the luminous intensity such as formula (70) of the backscatter light of 6th polarization image:
Wherein, p is the degree of polarization of the backscatter light of final underwater picture,For the 5th polarization image given zone
The luminous intensity of the backscatter light in domain,For the luminous intensity of the backscatter light of the 6th polarization image specific region.It should
Specific region is not have target and the uniform one piece of region of backscatter light in image.
Step 5.4, establish final underwater picture target emanation light luminous intensity computation model；
By the luminous intensity of the 5th polarization image, the luminous intensity of the 6th polarization image, final underwater picture total light intensity degree,
The degree of polarization of the backscatter light of final underwater picture, the luminous intensity of the backscatter light of final underwater picture, final underwater figure
The luminous intensity of infinite point backscatter light establishes the computation model of the luminous intensity of final underwater picture target emanation light as in.I.e.
According to formula (65), formula (66), formula (67), formula (68), the available final underwater picture target emanation of formula (70)
The computation model of the luminous intensity of light, shown in computation model such as formula (71):
Wherein, IObjectIt (x) is the luminous intensity of final underwater picture target emanation light, IToatlIt (x) is final underwater picture
Total light intensity degree, I'maxIt (x) is the luminous intensity of the 5th polarization image, I'minIt (x) is the luminous intensity of the 6th polarization image, p is final
The degree of polarization of the backscatter light of underwater picture, IBIt (x) is the luminous intensity of the backscatter light of final underwater picture,For
The luminous intensity of infinite point backscatter light in final underwater picture.
Formula (17) is the model for the clear scene finally established, the final underwater picture removal back obtained by the model
Scape scatters light, rebuilds the clear underwater scene of achromatization distortion.
The reason of the present embodiment is by research underwater picture cross-color, goes out from the transmission characteristic of underwater backscatter light
Hair establishes the light of the 5th polarization image based on depth information by the physical relation of analysis depth information of scene and scattering light
The model of the luminous intensity of the model of intensity and the 6th polarization image, so that the color information of restoration scenario, utilizes passive water later
Lower polarization imaging technology removes the backscatter light of final underwater picture, rebuilds the clearly final underwater figure of achromatization distortion
Picture.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of underwater polarization imaging method of natural light characterized by comprising
The first model of the third polarization image is established by the first polarization image and third polarization image；
The second model of the 4th polarization image, first polarization are established by the second polarization image and the 4th polarization image
Image and second polarization image are two orthogonal width polarization images of polarization state in an atmosphere, the third polarization image and
4th polarization image is the two width polarization images that polarization state is orthogonal under water；
The 5th polarization image is obtained according to the first model of the luminous intensity of the third polarization image and the third polarization image
Luminous intensity；
The 6th polarization image is obtained according to the second model of the luminous intensity of the 4th polarization image and the 4th polarization image
Luminous intensity；
Final underwater picture mesh is obtained according to the luminous intensity of the luminous intensity of the 5th polarization image and the 6th polarization image
Mark the luminous intensity of radiant light.
2. imaging method according to claim 1, which is characterized in that first polarization image includes the first red channel figure
As, the first green channel image and the first blue channel image, second polarization image includes the second red channel image, second green logical
Road image and the second blue channel image, the third polarization image include the red channel image of third, the green channel image of third and
Three blue channel images, the 4th polarization image include the 4th red channel image, the 4th green channel image and the 4th blue channel figure
Picture.
3. imaging method according to claim 2, which is characterized in that passing through the first polarization image and third polarization image
It establishes before the first model of the third polarization image, further includes:
Described first is obtained by the luminous intensity of the described first red channel image and the reflection constant of the first red channel image
The light source response rate of red channel image；
Described first is obtained by the luminous intensity of the described first green channel image and the reflection constant of the first green channel image
The light source response rate of green channel image；
Described first is obtained by the luminous intensity of the first blue channel image and the reflection constant of the first blue channel image
The light source response rate of blue channel image.
4. imaging method according to claim 2, which is characterized in that built by the first polarization image and third polarization image
Found the first model of the third polarization image, comprising:
Pass through the reflection constant of the described first red channel image, the light source response rate and the third of the first red channel image
The luminous intensity of red channel image establishes the first model of the red channel image of the third；
Pass through the reflection constant of the described first green channel image, the light source response rate and the third of the first green channel image
The luminous intensity of green channel image establishes the first model of the green channel image of the third；
Pass through the reflection constant of the first blue channel image, the light source response rate and the third of the first blue channel image
The luminous intensity of blue channel image establishes the first model of the third blue channel image.
5. imaging method according to claim 2, which is characterized in that passing through the second polarization image and the 4th polarization image
It establishes before the second model of the 4th polarization image, further includes:
Described second is obtained by the luminous intensity of the described second red channel image and the reflection constant of the second red channel image
The light source response rate of red channel image；
Described second is obtained by the luminous intensity of the described second green channel image and the reflection constant of the second green channel image
The light source response rate of green channel image；
Described second is obtained by the luminous intensity of the second blue channel image and the reflection constant of the second blue channel image
The light source response rate of blue channel image.
6. imaging method according to claim 2, which is characterized in that built by the second polarization image and the 4th polarization image
Found the second model of the 4th polarization image, comprising:
Pass through the reflection constant of the described second red channel image, the light source response rate and the described 4th of the second red channel image
The luminous intensity of red channel image establishes the second model of the 4th red channel image；
Pass through the reflection constant of the described second green channel image, the light source response rate and the described 4th of the second green channel image
The luminous intensity of green channel image establishes the second model of the 4th green channel image；
Pass through the reflection constant of the second blue channel image, the light source response rate of the second blue channel image and the described 4th
The luminous intensity of blue channel image establishes the second model of the 4th blue channel image.
7. imaging method according to claim 2, which is characterized in that according to the luminous intensity of the third polarization image and institute
The first model for stating third polarization image obtains the luminous intensity of the 5th polarization image, comprising:
It is red logical that the 5th is obtained according to the first model of the luminous intensity of the red channel image of the third and the red channel image of the third
The luminous intensity of road image；
It is green logical that the 5th is obtained according to the first model of the luminous intensity of the green channel image of the third and the green channel image of the third
The luminous intensity of road image；
It is blue logical that the 5th is obtained according to the first model of the luminous intensity of the third blue channel image and the third blue channel image
The luminous intensity of road image；
Pass through the luminous intensity of the 5th red channel image, the luminous intensity of the 5th green channel image and the 5th blue channel
The luminous intensity of image obtains the luminous intensity of the 5th polarization image.
8. imaging method according to claim 2, which is characterized in that according to the luminous intensity of the 4th polarization image and institute
The second model for stating the 4th polarization image obtains the luminous intensity of the 6th polarization image, comprising:
It is red logical that the 6th is obtained according to the first model of the luminous intensity of the 4th red channel image and the 4th red channel image
The luminous intensity of road image；
It is green logical that the 6th is obtained according to the first model of the luminous intensity of the 4th green channel image and the 4th green channel image
The luminous intensity of road image；
It is blue logical that the 6th is obtained according to the first model of the luminous intensity of the 4th blue channel image and the 4th blue channel image
The luminous intensity of road image；
Pass through the luminous intensity of the 6th red channel image, the luminous intensity of the 6th green channel image and the 6th blue channel
The luminous intensity of image obtains the luminous intensity of the 6th polarization image.
9. imaging method according to claim 1, which is characterized in that according to the luminous intensity of the 5th polarization image and institute
The luminous intensity for stating the 6th polarization image obtains the luminous intensity of final underwater picture target emanation light, comprising:
Obtain respectively the total light intensity degree of the final underwater picture, the backscatter light of the final underwater picture degree of polarization,
The light of infinite point backscatter light in the luminous intensity of the backscatter light of the final underwater picture, the final underwater picture
Intensity；
Pass through the luminous intensity of the 5th polarization image, the luminous intensity of the 6th polarization image, the final underwater picture
Total light intensity degree, the degree of polarization of the backscatter light of the final underwater picture, the final underwater picture backscatter light
The luminous intensity of infinite point backscatter light establishes the final underwater picture target spoke in luminous intensity, the final underwater picture
Penetrate the computation model of the luminous intensity of light；
Final underwater picture target emanation is obtained according to the computation model of the luminous intensity of the final underwater picture target emanation light
The luminous intensity of light.
10. imaging method according to claim 1, which is characterized in that the light of the final underwater picture target emanation light
The computation model of intensity are as follows:
Wherein, IObjectIt (x) is the luminous intensity of final underwater picture target emanation light, IToatlIt (x) is total light of final underwater picture
Intensity, I'maxIt (x) is the luminous intensity of the 5th polarization image, I'minIt (x) is the luminous intensity of the 6th polarization image, p is final underwater
The degree of polarization of the backscatter light of image, IBIt (x) is the luminous intensity of the backscatter light of final underwater picture,It is final
The luminous intensity of infinite point backscatter light in underwater picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810827464.8A CN109141638B (en) | 2018-07-25 | 2018-07-25 | A kind of underwater polarization imaging method of natural light |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810827464.8A CN109141638B (en) | 2018-07-25 | 2018-07-25 | A kind of underwater polarization imaging method of natural light |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109141638A CN109141638A (en) | 2019-01-04 |
| CN109141638B true CN109141638B (en) | 2019-11-26 |
Family
ID=64797851
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810827464.8A Active CN109141638B (en) | 2018-07-25 | 2018-07-25 | A kind of underwater polarization imaging method of natural light |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109141638B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111343368B (en) * | 2020-02-18 | 2021-08-20 | 清华大学 | Method and device for recovering depth of scattering medium based on polarization |
| CN112907472A (en) * | 2021-02-09 | 2021-06-04 | 大连海事大学 | Polarization underwater image optimization method based on scene depth information |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139347B (en) * | 2015-07-10 | 2018-12-14 | 中国科学院西安光学精密机械研究所 | A kind of polarization imaging defogging method of combination dark channel prior principle |
| CN105182362B (en) * | 2015-11-13 | 2017-06-27 | 北京航空航天大学 | A kind of natural water surface polarization remote sensing imaging simulation method |
| CN106407927B (en) * | 2016-09-12 | 2019-11-05 | 河海大学常州校区 | The significance visual method suitable for underwater target detection based on polarization imaging |
| CN107895348B (en) * | 2017-10-23 | 2021-09-14 | 天津大学 | Polarization image restoration method under non-uniform light field in scattering environment |
-
2018
- 2018-07-25 CN CN201810827464.8A patent/CN109141638B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109141638A (en) | 2019-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106485681B (en) | Underwater color image restoration method based on color correction and red channel prior | |
| Sahu et al. | A survey on underwater image enhancement techniques | |
| CN105761227B (en) | Underwater picture Enhancement Method based on dark channel prior and white balance | |
| CN105631829B (en) | Night haze image defogging method based on dark channel prior and color correction | |
| Wang et al. | Underwater image restoration via maximum attenuation identification | |
| US8836810B2 (en) | Imaging systems and methods for recovering object visibility | |
| CN109187364B (en) | High-concentration underwater polarization imaging method | |
| CN109141638B (en) | A kind of underwater polarization imaging method of natural light | |
| CN106981053A (en) | A kind of underwater picture Enhancement Method based on Weighted Fusion | |
| Zhou et al. | Single image dehazing motivated by Retinex theory | |
| CN107316278A (en) | A kind of underwater picture clearness processing method | |
| CN107403418A (en) | Defogging and the underwater picture Enhancement Method of color correction are carried out based on passage transmissivity | |
| WO2023070958A1 (en) | Image restoration method based on physical scattering model | |
| Benxing et al. | Underwater image recovery using structured light | |
| Zhang et al. | Atmospheric scattering-based multiple images fog removal | |
| Ma et al. | Single image defogging algorithm based on conditional generative adversarial network | |
| Fynbo et al. | The sources of extended continuum emission towards Q0151+ 048A: The host galaxy and the Damped Ly-alpha Absorber | |
| JP4945635B2 (en) | Saturated optical element | |
| CN109191441B (en) | Image processing method, apparatus, system and storage medium | |
| Shirkande et al. | Optimization of Underwater Image Enhancement Technique by Combining WCID and Wavelet Transformation Technique | |
| Moni et al. | Color balance and fusion for underwater image enhancement: Survey | |
| Hu et al. | Fast outdoor hazy image dehazing based on saturation and brightness | |
| Cecilia et al. | Visibility restoration of diverse turbid underwater images-two step approach | |
| CN117132752B (en) | Sand and dust image enhancement method, device, equipment and medium based on multidimensional weighting | |
| Gao et al. | Two‐Step Unsupervised Approach for Sand‐Dust Image Enhancement |
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 |