CN115439349A - Underwater SLAM optimization method based on image enhancement - Google Patents
Underwater SLAM optimization method based on image enhancement Download PDFInfo
- Publication number
- CN115439349A CN115439349A CN202210965506.0A CN202210965506A CN115439349A CN 115439349 A CN115439349 A CN 115439349A CN 202210965506 A CN202210965506 A CN 202210965506A CN 115439349 A CN115439349 A CN 115439349A
- Authority
- CN
- China
- Prior art keywords
- underwater
- image
- light
- optimization method
- frequency
- 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 31
- 238000005457 optimization Methods 0.000 title claims abstract description 21
- 238000012545 processing Methods 0.000 claims abstract description 11
- 230000000007 visual effect Effects 0.000 claims abstract description 11
- 238000000605 extraction Methods 0.000 claims abstract description 9
- 238000013507 mapping Methods 0.000 claims abstract description 6
- 230000009466 transformation Effects 0.000 claims abstract description 6
- 238000002834 transmittance Methods 0.000 claims abstract description 6
- 238000005286 illumination Methods 0.000 claims description 11
- 238000012634 optical imaging Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 5
- 238000013459 approach Methods 0.000 claims description 3
- 239000012736 aqueous medium Substances 0.000 claims description 3
- 230000014509 gene expression Effects 0.000 claims description 3
- 239000011521 glass Substances 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 239000011241 protective layer Substances 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 230000004807 localization Effects 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 238000003672 processing method Methods 0.000 description 3
- 238000000149 argon plasma sintering Methods 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Images
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
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/60—Extraction of image or video features relating to illumination properties, e.g. using a reflectance or lighting model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an underwater SLAM (simultaneous localization and mapping) optimization method based on image enhancement, which comprises the following steps of: s1, obtaining corrected feature point coordinates according to a refraction principle of light; s2, processing the gray value of each pixel point of the original image acquired by the underwater robot through linear transformation in the aspect of airspace; s3, a high-pass filter is used in the frequency domain, a proper cut-off frequency is set, and only signals above the cut-off frequency are allowed to pass; s4, introducing a dark primary color defogging algorithm on the land, and controlling the defogging degree, the minimum transmittance and the size of the partitioned window area; and S5, performing feature extraction in the traditional visual SALM on the image processed in the S4 to obtain a global map. The method improves the definition of the image acquired underwater, is beneficial to extracting better characteristic points and further obtains a clearer underwater three-dimensional model.
Description
Technical Field
The invention relates to the field of visual images, in particular to an underwater SLAM optimization method based on image enhancement.
Background
Currently, the visual SLAM technology is mainly used to describe that a robot synchronously performs mapping and self-positioning in an unknown place of an unknown environment. The robot establishes a map of an unknown environment through environment data acquired by a sensor, and then performs self positioning and map establishment by matching environmental features observed in motion with features in the map. The visual SLAM technology provides great help for robot positioning and navigation, and has the advantages of low cost, rich scene information and the like.
The existing visual SLAM technology is greatly influenced by the environment, so that most of the technology is applied to the land and the air in sunny weather. The underwater three-dimensional map reconstruction method is relatively few in use in an underwater environment, and is mainly caused by the fact that due to the particularity of the underwater environment, image information acquired by a vision sensor of the robot is refracted and scattered by light rays, and is influenced by environmental factors such as water quality and brightness, so that the image information is dim and fuzzy, subsequent feature extraction and matching are not facilitated, and the positioning accuracy and the final three-dimensional map reconstruction effect of the underwater robot are poor.
Therefore, how to design an image acquisition method by which a robot vision sensor can acquire high definition in an underwater environment is a problem to be solved by those skilled in the art.
Disclosure of Invention
The invention provides an underwater SLAM optimization method based on image enhancement by adopting an improved front-end image processing algorithm, so as to solve the problems.
The invention provides the following technical scheme:
an underwater SLAM optimization method based on image enhancement comprises the following steps:
s1, obtaining corrected feature point coordinates according to a refraction principle of light;
s2, processing the gray value of each pixel point of the original image acquired by the underwater robot through linear transformation in the aspect of airspace;
s3, a high-pass filter is used in the frequency domain, a proper cut-off frequency is set, and only signals above the cut-off frequency are allowed to pass;
s4, introducing a dark primary color defogging algorithm on the land, and controlling the defogging degree, the minimum transmittance and the size of the partitioned window area;
and S5, performing feature extraction in the traditional visual SALM on the image processed in the S4 to obtain a global map.
Preferably, in S1, the corrected feature point coordinate is obtained through a feature point coordinate equation, where the feature point coordinate equation is obtained as follows:
the point A is a projection point after underwater refraction, the point B is a projection point without underwater refraction, alpha and beta respectively represent incident angles of incident light and refracted light in the y-axis direction, theta and lambda respectively represent incident angles of the incident light and the refracted light in the x-axis direction, d is a distance between the protective layer glass and the lens, and coordinates of the corrected feature point (x is calculated through a feature point coordinate equation) B ，y B )。
Preferably, in S2, processing the gray value of each pixel point of the original image acquired by the underwater robot is represented as:
g(x,y)＝E H [f(x,y)]
where g (x, y) is the enhanced image, f (x, y) is the original image, E H Is an enhancement function.
Preferably, in S3, a time domain transform domain of the original image is converted into a fourier transform domain, a cut-off frequency is set, a high-pass filter is used to pass signals above the cut-off frequency, and signals below the cut-off frequency are filtered.
Preferably, in S4, the underwater optical imaging is modeled, and the underwater optical imaging model includes a reflection model and an illumination model, where expressions of the reflection model S and the illumination model E are as follows:
S＝L(x，y)R(x，y)
E(L(x，y)R(x，y))e -βd +E∞(1-e -βd )
wherein L is the illumination intensity of the incident light, and L (x, y) is at the coordinate position (x, y) where the light reaches the surface of the target object through the attenuation in the propagation process; r (x, y) represents a reflection function of a point on the target object at a coordinate position of (x, y); beta is the attenuation parameter of the aqueous medium; d is the light propagation length; e ∞ And E is the light intensity at distance and the light intensity reaching the underwater robot camera, respectively, where E is ∞ Is a constant.
Preferably, based on the underwater optical imaging model, the similarity of a land defogging algorithm, an underwater environment and a fog weather is integrated to obtain an underwater defogging function based on a dark primary color principle image, wherein the dark primary color principle image J is dark The definition is as follows:
in the formula, J c A certain color channel of J, r, g, b respectively represent three color channels; omega (x) is a region centered on x, J dark Is very low and approaches 0, J is the image acquired a priori, J dark The dark primary of J.
Preferably, assuming that the light factor is constant in a certain small area, it follows:
t(x)＝e -βd
wherein the content of the first and second substances,
expressing the light factor parameter of the atmosphere, E c (y) and S c (y) represents a certain color channel of the illumination model and the reflection model, respectively.
Preferably, by introducing a constant λ (0 < λ < 1), fog in a part of the scene is preserved by adjusting the value of λ:
wherein Ω (x) is a fixed matrix and t (x) represents a transmission factor; setting a lower limit t for the transmission factor t (x) 0 Calculating an image S (x) after the defogging process:
wherein E (x) is an atmospheric light component, E ∞ (x) Representing the maximum density pixel in the light component.
Preferably, in S5, the feature extraction step in the conventional visual SALM includes extracting ORB descriptors, feature matching, local mapping, loop detection, and global mapping.
The invention has the beneficial effects that:
the method is different from the traditional land vision SLAM image processing method, innovatively improves the original algorithm, and has the greatest advantages of solving the problems caused by the particularity of the underwater environment, such as dim and fuzzy images caused by light refraction, light scattering, water turbidity and the like, improving the definition of the images acquired underwater, being beneficial to extracting more excellent characteristic points and further obtaining a clearer underwater three-dimensional model.
Drawings
The invention is further illustrated by means of the attached drawings, but the embodiments in the drawings do not constitute any limitation to the invention, and for a person skilled in the art, without inventive effort, further drawings may be derived from the following figures.
FIG. 1 is a flowchart of an underwater SLAM optimization method based on image enhancement according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the refraction of rays by an underwater lens according to an embodiment of the present invention;
FIG. 3 is a schematic view of an underwater optical imaging model according to an embodiment of the present invention;
FIG. 4 is a flowchart of an underwater image processing algorithm with dark primary colors according to an embodiment of the present invention.
Detailed Description
An image enhancement-based underwater SLAM optimization method is described in further detail below with reference to specific embodiments, which are provided for purposes of comparison and explanation only, and the present invention is not limited to these embodiments.
Examples
Referring to fig. 1, the underwater SLAM optimization method based on image enhancement in the embodiment of the present invention includes the following steps:
s1, establishing a coordinate equation of the feature points before and after correction according to a refraction principle of light;
as shown in fig. 2, which is a principle diagram of refraction of rays of an underwater lens, point a is a refracted projection point, point B is an un-refracted projection point, angles of incidence a and β of incident light and refracted light in a y-axis direction are measured according to a real scene, angles of incidence θ and λ of incident light and refracted light in an x-axis direction are measured, and a new coordinate equation is established:
wherein d is the distance between the protective layer glass and the lens, and the coordinate (x) of the corrected characteristic point can be calculated B ，y B )。
S2, processing the gray value of each pixel point of the original image acquired by the underwater robot through linear transformation in the airspace aspect;
the spatial domain image enhancement is adopted, the gray value of the pixel point is directly processed, and the processing method can be expressed as follows:
g(x,y)＝E H [f(x,y)]
where g (x, y) is the enhanced image, f (x, y) is the original image, E H For enhancement functions, usually linear transformations, histogram equalization and histogram specification are used. Firstly, more balancing gray values which are distributed more intensively in the original image through histogram balancing of the original image; the histogram is specified through a histogram, a change function of a specified histogram is calculated, the equalized image histogram is subjected to linear transformation and is transformed into a corresponding specified histogram, and the contrast of the image can be obviously enhanced;
s3, a high-pass filter is used in the frequency domain, a proper cut-off frequency is set, and only signals above the cut-off frequency are allowed to pass;
the method has the advantages that the frequency domain image enhancement is adopted, the time domain transform domain of the original image is converted into the Fourier transform domain, a proper cut-off frequency is set, signals above the cut-off frequency are passed through a high-pass filter, signals below the cut-off frequency are filtered, the effect of enhancing the edge information of the image is achieved, the outline of the image information is clear, and the identification degree of the image is improved.
S4, introducing a dark primary color defogging algorithm on the land, and controlling the defogging degree, the minimum transmittance and the size of the partitioned window area;
FIG. 3 shows an underwater optical imaging model, where L is the illumination intensity of incident light, and L (x, y) is the coordinate position (x, y) of the surface of the target object reached through attenuation in the propagation process; r (x, y) represents a reflection function of a point of coordinate position (x, y) on the target object, the function being related only to the color characteristics of the target object itself; beta is the attenuation parameter of the aqueous medium; d is the light propagation length; e ∞ And E are the remote light intensity and the light intensity reaching the camera, respectively, where E ∞ Taking a constant. The expressions of the reflection model and the illumination model are as follows:
S＝L(x,y)R(x,y)
E(L(x,y)R(x,y))e -βd +E∞(1-e -βd )
according to the model, a land defogging algorithm is combined, and then the underwater defogging function can be obtained according to the similarity of the underwater environment and the foggy weather. Wherein the dark primary principle image is defined as:
in the formula, J c A certain color channel of J; Ω (x) is a region centered on x. Experiments prove that J dark The value of (d) is very low and approaches 0. Let J be the image acquired a priori, then J dark J, and the empirical rule obtained by observation experiments is called the dark channel prior rule. When image enhancement is performed by this method, the transmittance needs to be calculated. Here, the
The optical factor parameter representing the atmosphere, assuming that the optical factor is constant in a certain small area, can be further obtained:
t(X)＝e -βd
as known from the prior law of the dark primaries,
toward 0, then:
E c (y) and S c (y) represents a certain color channel of the illumination model and the reflection model, respectively.
In practice, even clear water with a background may contain some foreign particles and a distant scene may still be perceived as a haze. If the fog in the image is removed too thoroughly, depth information is lost. Therefore, at the time of defogging processing, a constant λ (0 < λ < 1) is introduced based on the above equation, and "fog" in a part of the scene is retained by adjusting the value of λ:
in the above formula, Ω (x) is taken as a fixed matrix, and therefore, t (x) obtained by the defogging treatment has a blocking effect, and the removal of the blocking effect makes the transmittance more accurate. In addition, when t (x) tends to zero, t (x) J (x) in the formula also tends to zero. Therefore, we set a lower limit t for the transmission factor t (x) 0 From this, the image after defogging processing can be calculated:
wherein E (x) is an atmospheric light component, E ∞ (x) Representing the maximum density pixel in the light component.
The dark primary underwater image processing algorithm flow is shown in fig. 4.
S5, performing feature extraction in the traditional visual SALM on the image processed in the S4 to obtain a global map;
and performing feature extraction in the traditional visual SALM on the processed image, extracting an ORB descriptor, performing feature matching, and removing points which do not meet the graph optimization by utilizing a PNP optimization matching pair. Estimating the relative motion of the current frame and the previous key frame according to the matched characteristic points, calculating the pose relationship, repositioning when tracking is lost, and tracking a local map for local optimization when enough matching time is available.
And then, local map building is carried out, the key frames are inserted into the map, the common view is updated, map points which do not meet the requirements are removed, new map points are triangulated, more matching pairs are searched in the domain key frames, local BA is executed, and then redundant local key frames are checked and removed.
In the loop detection step, whether a key frame exists in the queue is checked, in an embodiment, if the distance from the last loop is less than 10 key frames, loop detection is not needed, and the loop detection mainly performs appearance verification on the key frames and performs geometric verification on the relative motion relationship.
And finally, carrying out global map building, processing map points obtained in the previous step, starting a global BA, updating the poses of all key frames, updating all map points, and splicing into a final global map.
The embodiment of the invention mainly solves the problems caused by the particularity of the underwater environment, so that the underwater robot can well extract the characteristics in the motion process, and a clearer underwater three-dimensional model is constructed. The method is different from the traditional land vision SLAM image processing method, innovatively improves the original algorithm, and has the greatest advantages of solving the problems caused by the particularity of the underwater environment, such as dim and fuzzy images caused by light refraction, light scattering, water turbidity and the like, improving the definition of the images acquired underwater, being beneficial to extracting more excellent characteristic points and further obtaining a clearer underwater three-dimensional model.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (9)
1. An underwater SLAM optimization method based on image enhancement is characterized by comprising the following steps:
s1, obtaining corrected feature point coordinates according to a refraction principle of light;
s2, processing the gray value of each pixel point of the original image acquired by the underwater robot through linear transformation in the aspect of airspace;
s3, a high-pass filter is used in the frequency domain, a proper cut-off frequency is set, and only signals above the cut-off frequency are allowed to pass;
s4, introducing a dark primary color defogging algorithm on the land, and controlling the defogging degree, the minimum transmittance and the size of the partitioned window area;
and S5, performing feature extraction in the traditional visual SALM on the image processed in the S4 to obtain a global map.
2. The image enhancement-based underwater SLAM optimization method of claim 1, wherein in S1, the corrected feature point coordinates are obtained through a feature point coordinate equation, wherein the feature point coordinate equation is obtained as follows:
the point A is a projection point after underwater refraction, the point B is a projection point without underwater refraction, alpha and beta respectively represent incident angles of incident light and refracted light in the y-axis direction, theta and lambda respectively represent incident angles of the incident light and the refracted light in the x-axis direction, d is a distance between the protective layer glass and the lens, and coordinates (x is the coordinate of the corrected characteristic point) of the corrected characteristic point can be calculated through a characteristic point coordinate equation B ，y B )。
3. The image enhancement-based underwater SLAM optimization method according to claim 1, wherein in S2, the gray scale value of each pixel point of the original image acquired by the underwater robot is represented by:
g(x,y)＝E H [f(x,y)]
where g (x, y) is the enhanced image, f (x, y) is the original image, E H Is an enhancement function.
4. The image enhancement-based underwater SLAM optimization method according to claim 1, wherein in S3, a time domain transform domain of an original image is converted into a Fourier transform domain, a cut-off frequency is set, a high-pass filter is used for passing signals above the cut-off frequency, and signals below the cut-off frequency are filtered.
5. The image enhancement-based underwater SLAM optimization method according to claim 1, wherein in S4, underwater optical imaging is modeled, the underwater optical imaging model comprises a reflection model and an illumination model, wherein the expressions of the reflection model S and the illumination model E are as follows:
S＝L(x,y)R(x,y)
E＝(L(x,y)R(x,y))e -βd +E ∞ (1-e -βd )
wherein L is the illumination intensity of the incident light, and L (x, y) is at the coordinate position (x, y) where the light reaches the surface of the target object through the attenuation in the propagation process; r (x, y) represents a reflection function of a point on the target object at coordinate position (x, y); beta is the attenuation parameter of the aqueous medium; d is the light propagation length; e ∞ And E is the light intensity at distance and the light intensity reaching the underwater robot camera, respectively, where E is ∞ Is a constant.
6. The image enhancement-based underwater SLAM optimization method of claim 5, wherein an underwater defogging function based on a dark primary color principle image J is obtained by integrating the similarity of a land defogging algorithm, an underwater environment and a foggy weather based on the underwater optical imaging model, wherein the dark primary color principle image J is a black primary color principle image dark The definition is as follows:
in the formula, J c A certain color channel of J, r, g, b respectively represent three color channels; omega (x) is a region centered on x, J dark Is very low and approaches 0, J is the image acquired a priori, J dark The dark primary of J.
7. The image enhancement based underwater SLAM optimization method of claim 6, wherein assuming that the light factor is constant in a certain small area, we derive:
t(x)＝e -βd
wherein, the first and the second end of the pipe are connected with each other,
parameter of light factor representing atmosphere, E c (y) and S c (y) represents a certain color channel of the illumination model and the reflection model, respectively.
8. The method of image enhancement based underwater SLAM optimization of claim 7, characterized in that fog in a part of the scene is preserved by introducing a constant λ (0 < λ < 1), by adjusting the value of λ:
wherein Ω (x) is a fixed matrix and t (x) represents a transmission factor; setting a lower limit t for the transmission factor t (x) 0 Calculating an image S (x) after the defogging process:
wherein E (x) is an atmospheric light component, E ∞ (x) Representing the maximum density pixel in the light component.
9. The image enhancement-based underwater SLAM optimization method of claim 1, wherein in S5, the feature extraction steps in the traditional visual SALM comprise ORB descriptor extraction, feature matching, local mapping, loop detection and global mapping.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210965506.0A CN115439349A (en) | 2022-08-12 | 2022-08-12 | Underwater SLAM optimization method based on image enhancement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210965506.0A CN115439349A (en) | 2022-08-12 | 2022-08-12 | Underwater SLAM optimization method based on image enhancement |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115439349A true CN115439349A (en) | 2022-12-06 |
Family
ID=84242136
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210965506.0A Pending CN115439349A (en) | 2022-08-12 | 2022-08-12 | Underwater SLAM optimization method based on image enhancement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115439349A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116580291A (en) * | 2023-05-24 | 2023-08-11 | 中国海洋大学 | Visual synchronous positioning and mapping method and system for underwater turbid strong scattering |
-
2022
- 2022-08-12 CN CN202210965506.0A patent/CN115439349A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116580291A (en) * | 2023-05-24 | 2023-08-11 | 中国海洋大学 | Visual synchronous positioning and mapping method and system for underwater turbid strong scattering |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108596849B (en) | Single image defogging method based on sky region segmentation | |
| CN110232666B (en) | Underground pipeline image rapid defogging method based on dark channel prior | |
| CN111079556A (en) | Multi-temporal unmanned aerial vehicle video image change area detection and classification method | |
| Tripathi et al. | Removal of fog from images: A review | |
| CN110956661B (en) | Method for calculating dynamic pose of visible light and infrared camera based on bidirectional homography matrix | |
| JP5982719B2 (en) | Image fog removal apparatus, image fog removal method, and image processing system | |
| CN112837233A (en) | Polarization image defogging method for acquiring transmissivity based on differential polarization | |
| TWI489416B (en) | Image recovery method | |
| CN111553862B (en) | Defogging and binocular stereoscopic vision positioning method for sea and sky background image | |
| CN110675340A (en) | Single image defogging method and medium based on improved non-local prior | |
| CN110245600B (en) | Unmanned aerial vehicle road detection method for self-adaptive initial quick stroke width | |
| CN111210396A (en) | Multispectral polarization image defogging method combined with sky light polarization model | |
| CN111833258A (en) | Image color correction method based on double-transmittance underwater imaging model | |
| CN112465708A (en) | Improved image defogging method based on dark channel | |
| CN112561996A (en) | Target detection method in autonomous underwater robot recovery docking | |
| CN111539246A (en) | Cross-spectrum face recognition method and device, electronic equipment and storage medium thereof | |
| Bansal et al. | A review of image restoration based image defogging algorithms | |
| CN113610728A (en) | Polarization double-image defogging method based on four-dark-channel mean comparison | |
| CN115439349A (en) | Underwater SLAM optimization method based on image enhancement | |
| CN111951339A (en) | Image processing method for performing parallax calculation by using heterogeneous binocular cameras | |
| CN113487509B (en) | Remote sensing image fog removal method based on pixel clustering and transmissivity fusion | |
| CN111899269B (en) | Unmanned aerial vehicle image and SAR satellite image matching method based on edge structure information | |
| CN112862721A (en) | Underground pipeline image defogging method based on dark channel and Retinex | |
| CN109191405B (en) | Aerial image defogging algorithm based on transmittance global estimation | |
| CN116757949A (en) | Atmosphere-ocean scattering environment degradation image restoration method and system |
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 |