CN110415178A - A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light - Google Patents
A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light Download PDFInfo
- Publication number
- CN110415178A CN110415178A CN201910480925.3A CN201910480925A CN110415178A CN 110415178 A CN110415178 A CN 110415178A CN 201910480925 A CN201910480925 A CN 201910480925A CN 110415178 A CN110415178 A CN 110415178A
- Authority
- CN
- China
- Prior art keywords
- channel
- image
- energy
- intensity
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000005352 clarification Methods 0.000 title abstract 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 71
- 230000005540 biological transmission Effects 0.000 claims abstract description 30
- 238000003384 imaging method Methods 0.000 claims description 28
- 238000002834 transmittance Methods 0.000 claims description 21
- 238000003707 image sharpening Methods 0.000 claims description 11
- 238000005286 illumination Methods 0.000 claims description 6
- 230000035515 penetration Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 21
- 238000011084 recovery Methods 0.000 abstract description 4
- 238000007796 conventional method Methods 0.000 abstract description 2
- 230000015572 biosynthetic process Effects 0.000 abstract 1
- 238000010521 absorption reaction Methods 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 9
- 238000011156 evaluation Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000002238 attenuated effect Effects 0.000 description 4
- 239000002245 particle Substances 0.000 description 4
- 238000000149 argon plasma sintering Methods 0.000 description 3
- 230000015556 catabolic process Effects 0.000 description 3
- 238000006731 degradation reaction Methods 0.000 description 3
- 230000014509 gene expression Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 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 1
- 238000012937 correction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 239000013535 sea water Substances 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The present invention provides a kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light.The present invention estimates the scattering coefficient in other channels according to wavelength-scattering coefficient linear relation and calculates its transmissivity, is accurately estimated according to bias light formation basic theory water body bias light.For this serious problem of colour cast, the present invention determines target and water distance between the surface D according to electromagnetic wave dump energy ratio and least square method, estimates energy attenuation amount of the water surface at depth of water D with this, realizes the compensation to restored image color.The method of the present invention can effectively estimate transmission coefficient and the energy attenuation of bias light and each channel, realize and restore to image.The experimental results showed that recovery effect of the present invention is fairly obvious, color rendition degree is very high, and effect is better than other conventional methods.
Description
Technical Field
The invention relates to the technical field of image restoration, in particular to an underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation.
Background
Scatter and color distortion are two major sources of underwater photographic degradation. Light scattering is due to multiple reflections of light rays by particles in the water before they strike the camera. The visibility and contrast of the captured image is reduced due to the scattering effect. The color distortion corresponds to the attenuation difference of light rays with different wavelengths when the light rays are transmitted underwater, and R, G, B changes caused by the difference of energy residue of three channels. The most severe attenuation of the R channel energy and the least attenuation of the B channel energy makes the presentation appear to show that the surrounding underwater environment is predominantly blue in color. At present, almost all underwater processing technologies cannot recover underwater images very effectively.
Traditionally, processing of underwater images has focused only on removing light scatter or color adjustments. The underwater image processing algorithm for removing the light scattering distortion comprises the steps of recovering the definition of an underwater image based on the defogging of the underwater image, removing a blurring effect by combining a point spread function and a modulation transfer function, and compensating the image with reduced visibility by using a polarization effect. Although the above method enhances visibility with respect to an image, distortion due to a wavelength attenuation difference, i.e., a color change, still exists. On the other hand, algorithms for color change correction estimate underwater environmental parameters generally by color registration taking into account light attenuation, equalize the brightness distribution of the colors using equalization of the RGB and HSI color spaces, and dynamically mix illumination of objects in a distance-dependent manner using controllable polychromatic light sources to compensate for color loss. Although the color balance is improved, these methods are effective for eliminating the image blurring effect caused by light scattering. Therefore, both the scattering and the difference in the energy attenuation of the color channels at different wavelengths need to be considered. Aiming at the scattering and attenuation phenomena generated when electromagnetic waves are transmitted underwater, an underwater imaging model is required to be further refined, and therefore a restored image with higher quality can be obtained. In the prior art, the restoration of an underwater image is mainly based on the following principle:
1. underwater imaging model
The Jaffe-McGlamry model assumes that the target is an ideal Lambertian, and estimation of the direct illumination light component can be achieved by geometric optics. The forward scatter component is the reflected light that reaches the imaging device from small angle scattering and can be calculated by the point spread function convolution operation described above. Backscattered light is reflected light that reaches the imaging device by large angle scattering. The model shows that the distance between the imaging device and the target object can affect the quality of the acquired underwater image. The effect of the scattering effect will be greater with increasing distance. From the Jaff-McGlamry model, in an underwater environment, the light intensity received by the camera can be expressed as direct light JdForward scattered light JfBackward scattered light Jb. The total irradiance I reaching the camera is a linear superposition of these three parts, which can be represented by the following equation:
I＝Jd+Jf+Jb (13)
1.1 direct light
The light rays received by the imaging device and reflected by the target are called refraction light, and the expression is as follows:
wherein J (x, λ) represents the intensity of light reflected by the target object, Jd(x, λ) represents the light intensity of the target reflected light after attenuation and received by the imaging device, αλAnd betaλRespectively representing an absorption coefficient and a scattering coefficient, wherein lambda represents wavelength and corresponds to three channels of RGB of an image, because the water body has different absorption degrees for light with different wavelengths, the attenuation coefficient can change along with the change of the wavelength of the channel, and d (x) represents the distance between an object and a camera.
1.2 Back-scattered light
The backscattered light does not contain target reflected light, is light entering the camera after ambient light is scattered by objects (such as suspended particles) in water, and has the expression:
where B (λ) is referred to as background light.
Neglecting the effect of forward scattering, the total light intensity can be expressed as:
wherein,representing the transmission coefficient of attenuation of light waves due to scattering,representing the transmission coefficient of attenuation of light waves due to water absorption. In some methods tα(x) Considered as a constant, denoted by K. In fact tα(x) Distance tod (x) is related to the wavelength λ, and tα(x) Considering a constant less reasonable, the invention proposes a new tα(x) And (6) solving a method.
The attenuation of the B channel due to water absorption is small relative to the R channel, so the present invention ignores the attenuation of the B channel by water absorption. The imaging model for the B channel is defined here as:
I(x,λ)＝J(x,λ)tB(x)+B(λ)[1-tB(x)],λ∈(B) (17)
wherein, tB(x) Representing the transmission coefficient of the B channel.
The imaging model for the G channel is:
I(x,λ)＝J(x,λ)tG(x)+B(λ)[1-tG(x)],λ∈(G) (18)
wherein, tG(x) Representing the transmission coefficient of the B channel.
The imaging model for the R channel is:
I(x,λ)＝J(x,λ)tR(x)+B(λ)[1-tR(x)],λ∈(R) (19)
wherein, tR(x) Representing the transmission coefficient of the R channel.
The light intensity I (x, λ) received by the camera is the acquired underwater image, representing a blurred image. As can be seen from expressions (17), (18) and (19), in order to restore a clear image J (x, λ), the transmission coefficient t for each channel is requiredB(x),tG(x)tR(x) And background light B (λ).
Disclosure of Invention
The invention aims to provide an underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation aiming at the technical defects in the prior art, so as to solve the technical problem that the conventional method in the prior art is poor in image restoration quality.
The invention also aims to solve the technical problem of how to improve the contrast and definition of an image and improve color distortion in the process of underwater image restoration.
In order to achieve the technical purpose, the invention adopts the following technical scheme:
an underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation is characterized in that each channel of an image is restored according to the following formula (1); meanwhile, the energy of the image is compensated according to the following formula (2):
in the formula (1), J (x, lambda) represents the energy intensity of each channel of RGB after restoration, I (x, lambda) represents the energy intensity of each channel of RGB of the original image, B (lambda) represents the light intensity of the background light of the water body, tλ(x) Representing the transmission coefficients of the respective channels of RGB.
In the formula (2), Nrer (λ) represents the normalized residual energy ratio, D represents the distance between the target and the water surface, J (x, λ) represents the energy intensity of each channel of RGB after restoration, and J (x, λ) represents the energy intensity of each channel of RGB after restorationKAnd (x, lambda) represents the energy intensity of each channel of the final restored image RGB after compensating the energy attenuation from the water surface to the depth D.
Preferably, the light intensity B (λ) of the background light of the water body is obtained by the following method:
when underwater imaging is carried out, after entering underwater, atmospheric light is scattered when encountering suspended impurities in a water body in the transmission process and then enters a camera from all directions, namely the water body background light. Therefore, the backscattering component in natural environment can be expressed by the following formula:
wherein,theta represents the angle of scattering,angle between reflecting surface representing object and image forming apparatus]. En denotes the target positionThe intensity of the light. c represents the attenuation coefficient of the light during transmission. F represents the focal length of the camera and dx is the distance between the target and the imaging device. β (θ) represents the volume scattering function. In practice, F is much larger than dx, so (1-F/dx) can be considered approximately as 1. Thus, equation (3) can be written as:
where k represents an empirical constant that can be considered approximately as 1.
Calculating the back-down backscattering for each direction, and then performing an integral calculation, we can obtain:
from the scattering coefficient calculation obtained from the above analysis, equation (5) can be further simplified:
finally, as can be seen from equation (6), we can obtain R, G, B channel water background light intensity by scattering coefficient and attenuation coefficient. Wherein E isnWe replace it with the pixel value with the highest brightness for each channel in the image. Next, the transmittance, attenuation coefficient, and scattering coefficient of each channel are calculated.
Preferably, the transmission coefficient t of the B channelB(x) Is obtained according to the following formula (7):
in the formula (7), tB(x) The transmission coefficients of the B channel are represented, B (λ) represents the intensity of the background light of the water body, I (x, λ) represents the intensity of the three color channels of the original image R, G, B, and Ω (y) represents the image area.
Preferably, the attenuation coefficient c of each channelλIs obtained according to the following formula (8):
where d (x) represents the distance between the target and the camera.
Preferably, the scattering coefficients of the G, R two channels are obtained according to the following equation (9):
bλ＝(1.62517-0.00113λ)ba (9)
preferably, G, R two-channel transmittance tG(x)、tR(x) Coefficient of attenuation cG、cRIs obtained according to the following formula (10):
preferably, the distance d (x) between the imaging device and the target is obtained according to the following formula (11):
wherein, tB(x) Represents the transmission coefficient of the B channel, and nrer (B) represents the normalized residual energy ratio.
Preferably, the normalized residual energy ratio nrer (b) is 0.95.
Preferably, the distance D between the target and the water surface is obtained according to the following equation (12):
wherein, the value range of the lambda belongs to { R, G, B }.Representing the amount of energy per color channel when natural light reaches the water surface,representing the remaining energy of each color channel after penetration of the water to a certain depth, Nrer (lambda)DIs the ratio of the remaining energies at different wavelengths.
The invention provides an underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation. And acquiring the transmissivity and the scattering coefficient of the B channel by a DCP theory and an underwater imaging model. And estimating R, G the scattering coefficient and the transmittance of the two channels according to the wavelength-scattering coefficient ratio formula. And estimating the background light intensity of the water body by using a water body background light intensity calculation formula, wherein the illumination intensity is represented by the maximum brightness value of each channel, and the distance d (x) between the target and the imaging equipment is estimated by using the transmittance of the B channel and the unit energy residual ratio. And obtaining a restored image J (x, lambda) by the underwater imaging model and the estimated water body background light intensity and transmittance of each channel. Calculating the distance D between the object and the water surface by using the unitized residual energy ratio and the least square method, calculating the energy attenuation of R, G, B three channels when the natural light reaches the depth D, and compensating the energy of the restored image to obtain the final clear image JK(x, λ). In order to verify the invention, the quality of the experimental result image is evaluated by subjective evaluation and objective evaluation. The evaluation result shows that compared with the traditional algorithm, the method can more effectively compensate and deblur the image color, so that the restored image is more real.
Drawings
FIG. 1 is a schematic illustration of an underwater optical imaging model;
FIG. 2 is a schematic flow diagram of the process of the present invention;
FIG. 3 is a graph showing the comparison of the effects of different image restoration methods in embodiment 2 of the present invention;
FIG. 4 is a graph showing a comparison of effects of different image restoration methods in embodiment 2 of the present invention;
FIG. 5 is a comparison chart of the effects of different image originals and the restoration method of the present invention in embodiment 2 of the present invention;
in FIGS. 3 to 4, (a) shows an original image; (b) represents an Image restored by the method disclosed in the document "He K, Sun J, Tang X. Single Image Haze Removal Using Dark Channel Prior [ J ]. IEEE Transactions on Pattern Analysis & Machine understanding, 2011,33(12): 2341-; (c) represents an image restored by a method disclosed in the document "Wen H, Tian Y, Huang T, et al.single underserver image enhancement with a new optical model [ C ]. IEEE International Symposium on Circuits and systems, IEEE,2013: 753-; (d) the image restored by the method of the present invention is shown.
Detailed Description
Hereinafter, specific embodiments of the present invention will be described in detail. Well-known structures or functions may not be described in detail in the following embodiments in order to avoid unnecessarily obscuring the details. Approximating language, as used herein in the following examples, may be applied to identify quantitative representations that could permissibly vary in number without resulting in a change in the basic function. Unless defined otherwise, technical and scientific terms used in the following examples have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Example 1 (Underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation)
1. Theory of DCP
McCarney constructs an atmospheric scattering model of light waves under a foggy condition:
I(x)＝J(x)t(x)+A[1-t(x)] (20)
wherein I (x) represents the image captured by the camera, J (x) represents the reflected light of the object and is the required fog-free image, A represents the light intensity of the atmosphere, t (x) represents the transmission coefficient of the light wave in the atmosphere and represents the degree of the reflected light of the object attenuated by the atmosphere, J (x) t (x) represents the light intensity after the attenuation of the light containing the information of the object, and A [1-t (x) ] represents the light intensity of the atmosphere received by the camera. (20) The formula reveals the reason of the image quality degradation in the foggy weather, and the image defogging is equivalent to solving J (x) from the formula (20) to obtain a clear image.
Dark primaries are typically in shadow, black, or bright-colored objects. So there may be pixels with very low brightness in the RGB three channels in each local area, and this statistical rule is called dark primary prior, and the dark primary itself has no brightness or very low brightness.
Wherein, Jc(x) Represents the intensity of each color channel of J (x), and Ω (y) represents a local region of the image, Jdark(x) The dark primary color of J (x).
According to the dark primary color theory, the dark primary color value can be approximated as the minimum brightness operation of the RGB channel of each pixel point in the foggy image, namely
Wherein c represents three channels of the original image,the minimum brightness value is obtained for the three color channels in the original image. To estimate the transmission, assuming that the atmospheric light intensity A is known, the minimum value is taken for equation (20)
According to the dark channel prior theory, the dark channel value of the fog-free image tends to zero, i.e. the dark channel value of the fog-free image tends to zero
Then (22) can be changed into
Dividing both sides of the formula (25) by A to change the formula (25) into
From this, the transmittance of the atmosphere can be determined:
the image after defogging can be obtained by the equation (20):
2. theoretical value calculation of water body background light
Due to the suspended particles in water, electromagnetic waves can generate a very strong backscattering effect when being transmitted underwater. This makes the underwater image we obtain appear very blurred and the image quality is degraded. In studies by numerous scholars, it has been found that the accuracy of the estimate of the intensity of the water body determines whether the algorithm can successfully recover degraded underwater images. One side will reach the camera after multiple scattering, which is why the underwater image usually obtained has 'fog', which causes the image quality to be seriously degraded. The water body light intensity estimation result of each channel R, G, B influences the image restoration, wherein the water body background light intensity of the R channel has the largest influence on the image restoration. B. The G two channels are relatively less influential.
For the above analysis, we can summarize that the estimation of the water body light intensity is crucial to the recovery effect of the underwater image, the proportion occupied by the water body light intensity of the R channel is the highest, the influence of the G channel is smaller, and the relative influence of the B channel is the smallest. The estimation of the water body light intensity will be performed below.
When underwater imaging is carried out, after entering underwater, atmospheric light is scattered when encountering suspended impurities in a water body in the transmission process and then enters a camera from all directions, namely the water body background light. Therefore, the backscattering component in natural environment can be expressed by the following formula:
wherein,theta represents the angle of scattering,representing the angle of the reflecting surface of the object with the imaging device. En represents the illumination intensity of the target location. c represents the attenuation coefficient of the light during transmission. F represents the focal length of the camera and dx is the distance between the target and the imaging device. β (θ) represents the volume scattering function. In practice, F is much larger than dx, so (1-F/dx) can be considered approximately as 1. Thus, equation (3) can be written as:
where k represents an empirical constant that can be considered approximately as 1.
Calculating the back-down backscattering for each direction, and then performing an integral calculation, we can obtain:
from the scattering coefficient calculation obtained from the above analysis, equation (5) can be further simplified:
finally, as can be seen from equation (6), we can obtain R, G, B channel water background light intensity by scattering coefficient and attenuation coefficient. Where En we replace with the highest luminance pixel value for each channel in the image. Next, the transmittance, attenuation coefficient, and scattering coefficient of each channel are calculated.
3. Estimating the transmission coefficient and attenuation coefficient of each channel
The scattering and absorption effects of a large amount of micro suspended particles such as minerals, plankton and the like in water cause the underwater shot image to show the characteristics of blurring, low contrast, color distortion and the like, and the detailed information contained in the image is blurred, so that the application of various underwater imaging systems is directly influenced. Because the degradation principle of the haze image is very similar to that of the underwater image, the good effect obtained by the DCP theory for enhancing the haze image is inspired, and a certain effect is obtained by applying the DCP to deblurring the underwater image. However, the research of the DCP on deblurring of underwater images is slightly insufficient, and the absorption degree of the water body to electromagnetic waves is related to the wavelength of the electromagnetic waves, so that the color distortion of the underwater images is caused. In an underwater environment, the longer the wavelength of an electromagnetic wave, the faster it decays. Water absorbs most red and orange light and least blue light. Moreover, as the propagation depth of the electromagnetic wave in the water body is increased, the energy proportion imbalance phenomenon of each channel is more obvious.
3.1 estimating the transmittance of the B channel
As analyzed above, for the problem of severe attenuation of the red channel, we can determine the dark channel image by using only G, B channels, as in the method proposed in chapter iii, which can avoid that the R channel is severely attenuated and the R channel is a dark channel image when the energy remaining is close to zero using the general DCP principle. But this method itself also means that the dark channel image is found to be less accurate. And when the attenuation of the energy of the R channel is not too serious, the phenomenon of overexposure of the recovered image also occurs. Therefore, when a dark channel image is obtained in this chapter, an R channel energy attenuation degree judgment formula is added.
We define that when the mean value of the pixel brightness in the acquired R channel of the underwater image is lower than 10, the energy attenuation of the R channel of the image is considered to be serious, and G, B two channels are used to acquire the dark channel image. Otherwise, we will use conventional DCP theory to solve for the dark primary channel. The decision formula can be expressed as:
where K represents the mean value of the brightness of all the pixels of the R channel, IR(x) Represents the x-th pixel of the R channel, and M multiplied by N represents the total number of pixels for acquiring the underwater image.
Because in the normally acquired underwater image, the energy attenuation amount of the B channel caused by the water body absorption is very small. Thus, the attenuation of the B channel under water is quite similar to the energy attenuation pattern when imaging in the presence of fog. Therefore, in this chapter, when the transmittance of each channel is obtained, the transmittance t of the B channel is estimated from the formula (7)B(x) In that respect 3.2 estimating the attenuation coefficient of the B channel
From the foregoing analysis of the underwater imaging model, it can be known that in underwater imaging, the transmittance is related to the relationship between the target and the camera, and is inversely proportional to the relationship between the target and the camera, and the transmittance gradually decreases as the distance increases. If the distance between the two is known, then the attenuation coefficient (approximated as the scattering coefficient) of the B channel can be calculated:
where d (x) represents the distance between the target and the camera, and the calculation method of d (x) will be described later.
We have now found the transmission coefficient t of the B channelB(x) In that respect The attenuation coefficient c of the B channel can be calculated by the formula (8)BThe attenuation of the B-channel energy is mainly due to scattering effects, so the scattering coefficient βBHere approximately equal to the attenuation coefficient cB。
3.3 estimating the transmittance and attenuation coefficient of the R, G channel
According to the relation between the wavelength and attenuation coefficient of the electromagnetic wave, the wavelength-scattering coefficient linear formula (9) of the electromagnetic wave is used, and according to the formula, the B-channel scattering coefficient beta is obtainedBAs the scattering coefficient of the reference wavelength, G, R two-channel scattering coefficient beta can be obtainedG，βR. As can be seen from the formula for calculating the transmittance, the ratio of the scattering coefficient of the G, R two channels to the scattering coefficient of the B channel allows G, R two-channel transmittance t to be calculated from the B channel transmittanceG(x)、tR(x) The formula is as follows:
after estimating G, R the transmission coefficients of the two channels according to the above formula, the attenuation coefficients of the two channels can be calculated by using the formula (8). The transmission coefficient and the attenuation coefficient are substituted into the formula (6), so that the water body background light intensity of R, G channels can be calculated, and the image is recovered by using the known quantities.
4 estimating the distance D (x) and the depth D
4.1 estimating distance based on electromagnetic wave residual energy ratio
In addition to being affected by the wavelength of electromagnetic waves, the unitized residual energy ratio Nrer (λ) is also affected by the salt content and concentration of phytoplankton. Thus, seawater is classified into types I, II and III. For each class I ocean unit distance traveled by the electromagnetic wave, the unitized residual energy ratios Nrer (λ) for red light (700um), green light (520um), and blue light (440um) are 82%, 95%, and 97.5%, respectively. The transmission coefficient t of the B channel is previously determinedB(x) Then the distance d (x) can be solved from equation (11):
the unitized residual energy ratio nrer (b) of the blue channel is taken to be 0.95.
So far, after calculating the distance d (x) between the target and the imaging device by the above formula, we can successfully calculate R, G, B the transmittance t of each of the three channelsB(x)、tG(x)、tR(x) And scattering coefficient betaB、βG、βRCoefficient of attenuation cB、cG、cRBackground light intensity of water body BB、BG、BR. From these acquired parameters, the underwater degraded image can be restored to obtain image J (x, λ). The recovery formula is:
through the above oneFrom the series of calculations we have a restored image j (x). We will now consider the next problem, for objects taken underwater, where the object will typically be at a vertical depth D from the surface. After natural light enters the water surface, the water body attenuates electromagnetic waves with different frequencies to different degrees, so that energy of R, G, B three channels is lost to different degrees after the natural light is transmitted for a distance D underwater. This will directly distort the acquired underwater image colors. Then, the work of the people is to estimate the distance D between the target and the water surface, calculate the energy residual ratio of the electromagnetic wave after the distance, and perform energy compensation on the image J (x, lambda) according to the ratio to obtain the final restored image JK(x,λ)。
4.2 distance D of target to surface of water
For natural illumination, when the light reaches the upper part of the water surface, the energy of the red, green and blue channels should be the same, i.e. the energy of the light reaches the upper part of the water surfaceIs uniform in energy level. After the water body penetrates through the water body to reach a certain depth, the residual energy attenuated by each color channel is the water body environment light intensity of each color channel. R, G, B the environmental light intensities of the water bodies of the three channels are respectivelyAndin order to estimate the depth D of the shooting target under water, we need to estimate R, G, B the residual energy of the three channels at the depth D. The water depth D is obtained using a least squares method on the attenuated form of the incident wave and the estimated residual energy. The formula is as follows:
wherein, the value range of the lambda belongs to { R, G, B }. The distance D between the target and the water surface can be determined. Using the remaining energy ratio Nrer (lambda) per wavelengthDTo compensate the energy lost by each channel, the color distortion problem can be solved:
wherein, the value range of the lambda belongs to { R, G, B }. J. the design is a squareK(x, λ) represents the restored image after compensating for the surface-to-depth D energy decay.
Example 2
This example is used to examine the practical effect of the method of example 1 on underwater image restoration.
In order to verify the effectiveness of the method, a plurality of images under different environments are selected for experiment and compared with the original image and the image processed by the other two methods disclosed by the literature. The experimental results are shown in FIGS. 3 to 5. From the experimental results, it can be seen that in different underwater environments, the b-group method directly applies the dark channel prior theory to underwater image processing, neglects the attenuation of light due to water absorption, and cannot correctly estimate the background light and the transmittance of the underwater environment, thereby causing the processing effect to be unobvious. Although the attenuation of light caused by water absorption is considered in the method of the group c, the transmission coefficient of each channel is not correctly estimated, so that the processed image has good deblurring effect, but the color is distorted. The method can correctly estimate the transmission coefficient of each channel, thereby well removing the blur of the underwater image and compensating the color, so that the restored image has obvious and clear details and natural color.
In order to realize a more accurate evaluation algorithm, the image is evaluated by using three image objective quality evaluation indexes of definition, contrast and color cast. The contrast C is characterized by a luminance component:
where M × N is the size of the image (sum of image pixels), and i (x) represents the gray-scale value at the x-position of the image.
The image definition is generally characterized by average gradients mg (mean gradients), which can well show the edge detail of the image and the variation degree between pixels of the image. The higher the value of MG, the more obvious the texture of the image is represented, and the higher the quality, MG is represented by the following formula:
wherein, Jλ(i, j) represents the pixel value size of a certain channel.
Color shift is related to the distribution characteristics of image chromaticity. The color deviation value K is defined as follows, the mean values of the color components a and b in the Lab color space are represented by mean _ a and mean _ b, and the calculation process of the value K is as follows:
d represents the average value of the color components a and b, and M _ a and M _ b are defined as follows:
m is defined as D, as follows:
color shift K can be expressed as:
K＝D/M (36)
the K value is proportional to the degree of color shift.
The evaluation results are shown in table 1.
TABLE 1 Objective evaluation results of the four methods
It can be seen from objective evaluation that the method of the invention not only improves the contrast and definition of the image, but also compensates the color distortion, and has satisfactory results in both subjective and objective evaluation, and the experimental results are superior to other methods.
Aiming at the influence of different channel energy attenuation degrees and the accuracy of water body background light estimation on a recovery result, the invention provides an underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation. The method can be used for more accurately estimating the background light of the water body on the basis of the background light forming principle. And estimating a dark channel by using the K value decision and calculating the transmissivity and the scattering rate of the B channel. The transmittance of the two channels was calculated G, R from the wavelength-scattering coefficient ratio. The distance between the target and the imaging device is estimated using the ratio of the residual energy of the underwater transmission of the electromagnetic waves. And estimating the distance between the target and the water surface by using a least square method, and calculating the energy loss of the electromagnetic wave from the water surface to the target position. And compensating the image according to the energy loss amount to obtain a final clear image. The objective evaluation result shows that the indexes of the method such as contrast, definition, color cast and the like are better than those of other methods.
The embodiments of the present invention have been described in detail, but the description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention. Any modification, equivalent replacement, and improvement made within the scope of the application of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. An underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation is characterized in that: performing restoration of an image according to the following formula (1); meanwhile, the image is compensated according to the following formula (2);
in the formula (1), J (x, lambda) represents the energy intensity of each channel of RGB after restoration, I (x, lambda) represents the energy intensity of each channel of RGB of the original image, B (lambda) represents the light intensity of the background light of the water body, tλ(x) Representing the transmission coefficients of the respective channels of RGB.
In the formula (2), Nrer (λ) represents the normalized residual energy ratio, D represents the distance between the target and the water surface, J (x, λ) represents the energy intensity of each channel of RGB after restoration, and J (x, λ) represents the energy intensity of each channel of RGB after restorationKAnd (x, lambda) represents the energy intensity of each channel of the final restored image RGB after compensating the energy attenuation from the water surface to the depth D.
2. The method of claim 1, wherein the intensity B (λ) of the background light of the water body is obtained according to the following formula (3), (4), (5) or (6):
wherein,theta represents the angle of scattering,reflecting surface and imaging device for representing targetAnd (4) preparing an included angle. En represents the illumination intensity of the target location. c represents the attenuation coefficient of the light during transmission. F represents the focal length of the camera and dx is the distance between the target and the imaging device. β (θ) represents the volume scattering function. In practice, F is much larger than dx, so (1-F/dx) can be considered approximately as 1. Where k represents an empirical constant that can be considered approximately as 1. EnWe replace it with the pixel value with the highest brightness for each channel in the image. Next, the transmittance, attenuation coefficient, and scattering coefficient of each channel are calculated.
3. The underwater image sharpening method based on electromagnetic wave energy residual ratio and background light estimation as recited in claim 1, wherein a transmission coefficient t of a B channelB(x) Is obtained according to the following formula (7):
in the formula (7), tB(x) The transmission coefficients of the B channel are represented, B (λ) represents the intensity of the background light of the water body, I (x, λ) represents the intensity of the three color channels of the original image R, G, B, and Ω (y) represents the image area.
4. The underwater image sharpening method based on the electromagnetic wave energy residual ratio and the background light estimation as recited in claim 1, wherein the attenuation coefficient c of each channelλIs obtained according to the following formula (8):
where d (x) represents the distance between the target and the camera.
5. The method of claim 1, wherein the scattering coefficients of G, R for two channels are obtained according to the following formula (9):
bλ＝(1.62517-0.00113λ)ba (9)
wherein, baIn the formula, λ represents a wavelength, and represents a scattering coefficient b corresponding to the wavelength λ, which is a scattering coefficient of a known wavelengthλ。
6. The underwater image sharpening method based on the electromagnetic wave energy residual ratio and the background light estimation as recited in claim 1, wherein the transmittance t of G, R two channelsG(x)、tR(x) Is obtained according to the following formula (14):
7. the method of claim 1, wherein the distance d (x) between the imaging device and the target is obtained according to the following formula (11):
wherein, tB(x) Represents the transmission coefficient of the B channel, and nrer (B) represents the normalized residual energy ratio. The normalized residual energy ratio nrer (b) is 0.95.
8. The method of claim 1, wherein the distance D between the target and the water surface is obtained according to the following equation (12):
wherein, the value range of the lambda belongs to { R, G, B }.Representing the amount of energy per color channel when natural light reaches the water surface,representing the remaining energy of each color channel after penetration of the water to a certain depth, Nrer (lambda)DIs the ratio of the remaining energies at different wavelengths.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910480925.3A CN110415178A (en) | 2019-06-06 | 2019-06-06 | A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910480925.3A CN110415178A (en) | 2019-06-06 | 2019-06-06 | A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110415178A true CN110415178A (en) | 2019-11-05 |
Family
ID=68358441
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910480925.3A Pending CN110415178A (en) | 2019-06-06 | 2019-06-06 | A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415178A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070683A (en) * | 2020-07-21 | 2020-12-11 | 西北工业大学 | Underwater polarization image restoration method based on polarization and wavelength attenuation joint optimization |
| CN112419179A (en) * | 2020-11-18 | 2021-02-26 | 北京字跳网络技术有限公司 | Method, device, equipment and computer readable medium for repairing image |
| CN114792294A (en) * | 2022-05-20 | 2022-07-26 | 陈恩依 | Underwater image color correction method based on attenuation coefficient |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW201308251A (en) * | 2011-08-04 | 2013-02-16 | Yi-Wu Chiang | Underwater image enhancement system |
| CN107507145A (en) * | 2017-08-25 | 2017-12-22 | 上海海洋大学 | A kind of underwater picture Enhancement Method based on the stretching of different colours spatially adaptive histogram |
| CN108564543A (en) * | 2018-04-11 | 2018-09-21 | 长春理工大学 | A kind of underwater picture color compensation method based on electromagnetic theory |
| CN108596853A (en) * | 2018-04-28 | 2018-09-28 | 上海海洋大学 | Underwater picture Enhancement Method based on bias light statistical model and transmission map optimization |
-
2019
- 2019-06-06 CN CN201910480925.3A patent/CN110415178A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TW201308251A (en) * | 2011-08-04 | 2013-02-16 | Yi-Wu Chiang | Underwater image enhancement system |
| CN107507145A (en) * | 2017-08-25 | 2017-12-22 | 上海海洋大学 | A kind of underwater picture Enhancement Method based on the stretching of different colours spatially adaptive histogram |
| CN108564543A (en) * | 2018-04-11 | 2018-09-21 | 长春理工大学 | A kind of underwater picture color compensation method based on electromagnetic theory |
| CN108596853A (en) * | 2018-04-28 | 2018-09-28 | 上海海洋大学 | Underwater picture Enhancement Method based on bias light statistical model and transmission map optimization |
Non-Patent Citations (1)
| Title |
|---|
| 谢昊伶: ""基于背景光估计与暗通道先验的水下图像复原"", 《光学学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112070683A (en) * | 2020-07-21 | 2020-12-11 | 西北工业大学 | Underwater polarization image restoration method based on polarization and wavelength attenuation joint optimization |
| CN112070683B (en) * | 2020-07-21 | 2024-03-12 | 西北工业大学 | Underwater polarized image restoration method based on polarization and wavelength attenuation combined optimization |
| CN112419179A (en) * | 2020-11-18 | 2021-02-26 | 北京字跳网络技术有限公司 | Method, device, equipment and computer readable medium for repairing image |
| CN114792294A (en) * | 2022-05-20 | 2022-07-26 | 陈恩依 | Underwater image color correction method based on attenuation coefficient |
| CN114792294B (en) * | 2022-05-20 | 2024-07-16 | 陈恩依 | Underwater image color correction method based on attenuation coefficient |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106991663B (en) | A kind of underwater colour-image reinforcing method theoretical based on dark | |
| Chiang et al. | Underwater image enhancement: using wavelength compensation and image dehazing (WCID) | |
| CN107507145B (en) | Underwater image enhancement method based on different color space self-adaptive histogram stretching | |
| Sahu et al. | A survey on underwater image enhancement techniques | |
| CN108564543A (en) | A kind of underwater picture color compensation method based on electromagnetic theory | |
| CN108765342A (en) | A kind of underwater image restoration method based on improvement dark | |
| CN102982514B (en) | The colour-image reinforcing method under water that a kind of dark primary and white balance combine | |
| CN110135434B (en) | Underwater image quality improvement method based on color line model | |
| CN110415178A (en) | A kind of underwater picture clarification method estimated based on electromagnetic wave energy residue ratio and bias light | |
| CN110322410B (en) | Underwater image defogging and color cast correction method based on bright channel transmissivity compensation | |
| CN104766284B (en) | The restored method of underwater colored blur image | |
| CN107527325B (en) | Monocular underwater vision enhancement method based on dark channel priority | |
| CN108257101B (en) | Underwater image enhancement method based on optimal recovery parameters | |
| Deng et al. | Underwater image enhancement based on removing light source color and dehazing | |
| CN111861896A (en) | UUV-oriented underwater image color compensation and recovery method | |
| TW201308251A (en) | Underwater image enhancement system | |
| Lu et al. | Single underwater image descattering and color correction | |
| CN110930319B (en) | Underwater image sharpening method based on self-adaptive transmissivity estimation | |
| Ghate et al. | New approach to underwater image dehazing using dark channel prior | |
| Jiwani et al. | Single image fog removal using depth estimation based on blur estimation | |
| CN110852971B (en) | Video defogging method based on dark channel prior and Retinex and storage medium | |
| CN115170404A (en) | Night image defogging method and device based on brightness compensation and layer separation | |
| Gujral et al. | A Novel Defogging Technique for Dehazing Images | |
| Zheng et al. | An illumination adaptive underwater image enhancement method | |
| Arora et al. | HOG and SIFT Transformation Algorithms for the Underwater Image Fusion |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191105 |
|
| WD01 | Invention patent application deemed withdrawn after publication |