CN110796674A - SAR image edge detection method and device combining wide line extraction - Google Patents
SAR image edge detection method and device combining wide line extraction Download PDFInfo
- Publication number
- CN110796674A CN110796674A CN201910976380.5A CN201910976380A CN110796674A CN 110796674 A CN110796674 A CN 110796674A CN 201910976380 A CN201910976380 A CN 201910976380A CN 110796674 A CN110796674 A CN 110796674A
- Authority
- CN
- China
- Prior art keywords
- edge intensity
- sar image
- edge
- preset
- wide line
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar image
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The application relates to an SAR image edge detection method and device combining wide line extraction. The method comprises the following steps: the method comprises the steps of obtaining edge intensity information of an SAR image in each direction, constructing a primary edge intensity image, carrying out convolution operation on the primary edge intensity image and a preset Gaussian kernel function to obtain a blackplug matrix corresponding to the primary edge intensity image, obtaining an enhanced edge intensity image of the SAR image according to a characteristic value and a characteristic vector of the blackplug matrix, filtering the enhanced edge intensity image by adopting a preset isotropic nonlinear filter to obtain wide lines in the enhanced edge intensity image, extracting the central line of the wide lines, and obtaining the edge of the SAR image. By adopting the method, the problems of weak edges and edge directions in SAR image edge detection can be solved simultaneously.
Description
Technical Field
The application relates to the technical field of synthetic aperture radar images, in particular to a method and a device for detecting SAR image edges by combining wide line extraction.
Background
SAR (Synthetic Aperture Radar) edge detection is an important component of SAR image interpretation, can provide auxiliary information for SAR image segmentation, feature extraction and target identification, and has attracted much attention in recent years. However, due to the diversity and complexity of the scene of the SAR image, the edges obtained from the SAR image are rough, the readability is poor, and weak edges are missing, so that the further processing result based on the edges of the SAR image is less than ideal. Therefore, in order to obtain ideal results through further processing such as SAR image segmentation, feature extraction, target recognition and the like, the performance of the SAR edge detector must be improved.
In the existing SAR image edge detection method, the most widely applied method is the edge detection method based on the ratio. The method can realize constant false alarm detection by calculating the ratio of the arithmetic mean value or the weighted mean value of a certain area of the image. Under ideal conditions, the method can accurately extract the step edges and has certain adaptability to the situation of multiple edges.
However, this type of method has weak edge detection capability and cannot obtain direction information of the edge, so that complete edge information cannot be provided for subsequent SAR image interpretation. How to solve the extraction of weak edges and edge directions in the edge detection of the SAR image is a technical problem to be solved urgently.
Disclosure of Invention
Based on this, it is necessary to provide a method, an apparatus, a computer device and a storage medium for detecting edges of SAR images, which can simultaneously solve the problem of weak edges and edge direction in the edge detection of SAR images.
A method for detecting edges of an SAR image by combining wide line extraction comprises the following steps:
acquiring edge intensity information of the SAR image in each direction, and constructing a preliminary edge intensity map;
performing convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a blackplug matrix corresponding to the preliminary edge intensity graph;
obtaining an enhanced edge intensity graph of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and extracting the central line of the wide line to obtain the image edge of the SAR image.
In one embodiment, the method further comprises the following steps: performing convolution filtering on each dimension of the SAR image according to a preset smoothing filter, and normalizing to obtain the edge strength of the SAR image in each direction; obtaining the final edge strength of each pixel point in the SAR image according to the edge strength of the SAR image in each direction; and constructing a preliminary edge intensity map according to the final edge intensity of each pixel point in the SAR image.
In one embodiment, the method further comprises the following steps: acquiring a preset wide line model, and acquiring a wide line structure in the preliminary edge intensity map according to the wide line model; performing convolution according to the wide line structure and a preset Gaussian kernel function to obtain each order partial derivative of each pixel point in the wide line structure; and constructing a black plug matrix corresponding to the preliminary edge intensity graph according to each order partial derivative of each pixel point.
In one embodiment, the method further comprises the following steps: acquiring a characteristic vector with the maximum absolute value in the black plug matrix as a preset direction vector; for each pixel point in the preliminary edge intensity graph, calculating the maximum absolute value of the second-order directional derivative of each pixel point in the direction of the directional vector; and obtaining an enhanced edge intensity map of the SAR image according to the maximum absolute value corresponding to each pixel point.
In one embodiment, the method further comprises the following steps: acquiring a preset isotropic nonlinear filter and a preset circular mask; classifying the pixel points to a gray level similarity weighting mask when the gray level value of the mask center of the circular mask and the gray level value of the pixel points in the enhanced edge intensity graph meet a preset similarity function by using the isotropic nonlinear filter; acquiring preset constant weight of the circular mask; comparing the mask center with other pixel points in the circular mask according to the constant weight to obtain a gray level similar weighting mask weight of the mask center; and carrying out inversion operation according to the gray level similarity weighting mask weight of the mask center to obtain a wide line in the enhanced edge intensity image.
In one embodiment, the method further comprises the following steps: and constructing the similarity function by taking the gray value of the center of the mask, the gray value of the pixel point and a preset threshold value as parameters in the hyperbolic tangent function.
In one embodiment, the method further comprises the following steps: constructing a step function according to the gray value of the center of the mask and the gray value of the pixel point; and constructing the similarity function by taking the step function and the threshold value as parameters in the hyperbolic tangent function.
A wide line extraction combined SAR image edge detection apparatus, the apparatus comprising:
the intensity map building module is used for obtaining the edge intensity information of the SAR image in each direction and building a preliminary edge intensity map;
the matrix acquisition module is used for carrying out convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a black plug matrix corresponding to the preliminary edge intensity graph;
the intensity enhancing module is used for obtaining an enhanced edge intensity image of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
the wide line extraction module is used for filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and the edge detection module is used for extracting the central line of the wide line to obtain the image edge of the SAR image.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring edge intensity information of the SAR image in each direction, and constructing a preliminary edge intensity map;
performing convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a blackplug matrix corresponding to the preliminary edge intensity graph;
obtaining an enhanced edge intensity graph of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and extracting the central line of the wide line to obtain the image edge of the SAR image.
A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, performs the steps of:
acquiring edge intensity information of the SAR image in each direction, and constructing a preliminary edge intensity map;
performing convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a blackplug matrix corresponding to the preliminary edge intensity graph;
obtaining an enhanced edge intensity graph of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and extracting the central line of the wide line to obtain the image edge of the SAR image.
The SAR image edge detection method combining the wide line extraction, the device, the computer equipment and the storage medium can construct a preliminary edge intensity image by acquiring the edge intensity information of the SAR image in all directions, the preliminary edge intensity image has no direction information, therefore, a mode of convolution by adopting a Gaussian kernel can be adopted, a black plug matrix of the preliminary edge intensity image is constructed, so that an enhanced edge intensity image of the SAR image can be obtained through the black plug matrix, the edge direction information of pixel points can be acquired through the black plug matrix, direction information is contained in the enhanced edge intensity image, finally, the wide line can be acquired through an isotropic nonlinear filter, the detection effect of weak edges can be improved, and meanwhile, the direction information of the edge pixel points is attached to a detection result.
Drawings
FIG. 1 is a flow chart of an SAR image edge detection method combining broad line extraction in one embodiment;
FIG. 2 is a schematic flow chart diagram illustrating the step of constructing a preliminary edge strength map in one embodiment;
FIG. 3 is a schematic diagram of the structure of a circular mask and wide lines in one embodiment;
FIG. 4 is a block diagram of an embodiment of an SAR image edge detection apparatus with wide line extraction combined;
FIG. 5 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In one embodiment, as shown in fig. 1, a method for detecting an edge of an SAR image by combining wide line extraction is provided, and the method can be applied to a terminal and includes the following steps:
and 102, acquiring edge intensity information of the SAR image in each direction, and constructing a preliminary edge intensity map.
The SAR image is an image acquired by the synthetic aperture radar, and the SAR image may be a 3D image or a 2D image, so that each direction refers to an X-axis direction and a Y-axis direction in the 2D image, and each direction refers to an X-axis direction, a Y-axis direction and a Z-axis direction in the 3D image.
The edge strength information refers to the strength of the edge of the SAR image, and the extraction of the edge strength information is beneficial to extracting the edge in the image and realizing the separation of the background and the edge.
The preliminary edge intensity map refers to an image obtained by describing pixel points by using edge intensity for each pixel point in the SAR.
And 104, performing convolution operation on the preliminary edge intensity map and a preset Gaussian kernel function to obtain a blackout matrix corresponding to the preliminary edge intensity map.
The gaussian kernel function may be selected from multiple orders of gaussian kernel functions as desired. The blackplug matrix is constructed from the various order partial derivatives of the preliminary edge intensity map.
And step 106, obtaining an enhanced edge intensity image of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix.
Specifically, the eigenvalue of the black matrix is related to the edge strength of the pixel point, and the eigenvector is related to the edge direction of the pixel point.
And step 108, filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph.
By filtering, the wide lines in the enhanced edge intensity map can be screened out.
And 110, extracting the central line of the wide line to obtain the image edge of the SAR image.
In the step, the central line of the wide line is used as the image edge of the SAR image.
In the SAR image edge detection method combining the wide line extraction, the initial edge intensity image can be constructed by acquiring the edge intensity information of the SAR image in each direction, and the initial edge intensity image has no direction information, therefore, a mode of convolution by adopting a Gaussian kernel can be adopted, a black plug matrix of the initial edge intensity image is constructed, and through the black plug matrix, an enhanced edge intensity image of the SAR image can be obtained, the edge direction information of pixel points can be acquired through the black plug matrix, so that the edge direction information is contained in the enhanced edge intensity image, finally, the wide line can be acquired through an isotropic nonlinear filter, the detection effect of weak edges can be improved, and meanwhile, the direction information of the edge pixel points is contained in the detection result.
In one embodiment, as shown in fig. 2, a schematic flow chart of the steps of constructing the preliminary edge strength map is provided, which includes the following steps:
and 202, performing convolution filtering on each dimension of the SAR image according to a preset smoothing filter, and normalizing to obtain the edge strength of the SAR image in each direction.
And 204, obtaining the final edge strength of each pixel point in the SAR image according to the edge strength of the SAR image in each direction.
And step 206, constructing a preliminary edge intensity map according to the final edge intensity of each pixel point in the SAR image.
In this embodiment, a smoothing filter is used for filtering, so that preliminary edge strength in the SAR map can be effectively extracted.
In a specific embodiment, the principle of the smoothing filter is based on linear minimum mean square error, the local mean in the detection window estimated by the filter is not an arithmetic mean, but an exponentially weighted mean, and the expression of the smoothing filter is:
f(x)＝Kexp(-α|x|)
where K denotes a normalization constant and α denotes a filter coefficient.
In the discrete case, f (n) can be passed through f1(n) and f2(n) the specific expression is as follows:
wherein N is 1,21(n) and f2The expression of (n) is:
wherein 0 < b ═ e-α< 1, a ═ 1-b, u (n) is the discrete Heaviside function.
Based on the principle of the smoothing filter, the SAR image is a two-dimensional image, the expression of the two-dimensional SAR image in the computer is I (x, y), the horizontal component at the edge intensity is first calculated, each column of the SAR image I (x, y) is convolved by using the smoothing filter f (y), and then the smoothing filter f is used1(x) And f2(x) Convolving each line of the SAR image I (x, y) to obtain
And
the expression is as follows:
wherein the content of the first and second substances,
representing the sign of the convolution operation in the horizontal direction,
indicating the sign of the convolution operation in the vertical direction.
The edge strength of I (x, y) along the horizontal direction can be determined by
Andthe normalization results are as follows:
similarly, a smoothing filter f (x) may be used to convolve each line of the SAR image I (x, y), and then smoothing may be usedWave filter f1(y) and f2(y) convolution is carried out on each line of the SAR image I (x, y), and the edge strength of the I (x, y) along the vertical direction edge can be obtained in the same way
After obtaining the horizontal edge strength and the vertical edge strength, the final edge strength can be obtained as follows:
from the final edge intensities, a preliminary edge intensity map can be obtained.
In one embodiment, the blackplug matrix may be calculated as follows: the method comprises the steps of obtaining a preset wide line model, obtaining a wide line structure in a preliminary edge intensity graph according to the wide line model, carrying out convolution on the wide line structure and a preset Gaussian kernel function to obtain each order partial derivative of each pixel point in the wide line structure, and constructing a black plug matrix corresponding to the preliminary edge intensity graph according to each order partial derivative of each pixel point.
Specifically, the wide line model is based on a wide line structure, the width is 2w, the height is an ideal line structure of h, and the expression is as follows:
to detect a line point in the wide line structure of the above formula, in a specific embodiment, for a line point m (x) in the wide line structure, a bright line m "(x) in a dark background is 0, and a dark line m" (x) > 0 in a bright background is used as a reference with the m "(x) amplitude value at m' (x) ═ 0.
To further suppress the effects of speckle, m' (x) and m "(x) should be estimated by the convolution of a gaussian smoothing kernel with m (x), which is:
thus, a complete scale-space description of the parabolic structure is obtained. The wide line point extraction method has the desirable characteristic that the position of the wide line point can be accurately detected, and meanwhile, the amplitude of the second derivative always takes the maximum value at the point, so that the remarkable wide line point can be detected based on the second derivative of the wide line structure.
In practical engineering, the parabolic structure can be analyzed in advance, so that only the analysis result of the parabolic structure is needed to be utilized in practical application.
Obtaining a wide line structure according to a preset wide line model, and performing convolution on the wide line structure and a Gaussian kernel function, wherein the expression is as follows:
wherein, r ″)b(x, σ, w, h) the condition for obtaining the maximum negative value of the amplitude at the center of the line structure is:
the two-dimensional wide line shows a one-dimensional curve structural characteristic in the direction perpendicular to the two-dimensional wide line, and the direction is defined as
The second directional derivative of that direction will take the absolute value maximum.
Then obtaining the wide line structure direction of each pixel point in the preliminary edge intensity graph, and defining the deviation of each order of the preliminary edge intensity graph as rx、ry、rxx、ryyAnd rxyThis can be obtained from the image with the following kernel functions:
thus, a blackplug Matrix (Hessian Matrix) can be constructed as:
in another embodiment, the step of obtaining the enhanced edge strength map is as follows: and for each pixel point in the preliminary edge intensity image, calculating the maximum absolute value of the second-order directional derivative of each pixel point in the direction of the directional vector, and obtaining an enhanced edge intensity image of the SAR image according to the maximum absolute value corresponding to each pixel point. In this embodiment, not only the edge strength can be calculated, but also the direction information of the edge can be obtained.
In one embodiment, the step of obtaining the wide line in the enhanced edge strength map comprises: the method comprises the steps of obtaining a preset isotropic nonlinear filter and a preset circular mask, classifying pixel points to a gray similar weighted mask when gray values of the center of the mask of the circular mask and gray values of pixel points in an enhanced edge intensity graph meet preset similar functions by the isotropic nonlinear filter, obtaining constant weight of the preset circular mask, comparing the center of the mask with other pixel points in the circular mask according to the constant weight to obtain gray similar weighted mask weight of the center of the mask, and performing inversion operation according to the gray similar weighted mask weight of the center of the mask to obtain wide lines in the enhanced edge intensity graph. In this embodiment, the wide lines of the enhanced edge intensity map can be effectively extracted by using the circular mask and the isotropic nonlinear filter.
In a specific embodiment, a schematic diagram of a circular mask and wide lines is shown in fig. 3, and the similarity function can be set as:
(x0,y0) The coordinates of the center point of the mask are taken as (x, y) coordinates of any point except the center point of the mask, I (x, y) is the gray value of the point (x, y), and the threshold value of t gray value comparison is adopted, and s in the mask is summed to obtain the final productObtain the weight of weighted mask Weighting Mask (WMSB).
Setting the circular mask as a constant weight, wherein the expression of the constant weight is as follows:
wherein r is the radius of the circular mask.
Normalizing the above equation can result in:
when the image is locally processed, the current pixel point is positioned in the center of the circular mask, and the pixel points outside the central point in the mask are compared with the central point, so that the following results are obtained:
c(x,y,x0,y0)＝ω0(x,y,x0,y0,r)×s(x,y,x0,y0,t)
c is the weight comparison result, so the WMSB weight at the mask center is:
for wide line response L, it can be derived from the inversion of WMSB weights:
wherein g is a geometric threshold, and g ═ mmax/2，mmaxIs the maximum value of m.
In another embodiment, the gray value of the center of the mask, the gray value of the pixel point and a preset threshold value can be used as parameters in the hyperbolic tangent function to construct the similarity function. In this embodiment, a hyperbolic tangent function is introduced to avoid a drastic change in s when a small change occurs near a threshold value, and an expression of a similarity function is as follows:
wherein sech (x) is 2/(e)x+e-x)。
In another embodiment, a step function can be constructed according to the gray value of the center of the mask, the gray value of the pixel point and the threshold value; and constructing a similar function by taking the step function as a parameter in the hyperbolic tangent function. In this embodiment, constructing a step function can extract bright lines or dark lines according to actual requirements, and is more suitable for actual engineering.
Specifically, the step function is set as follows:
then the similarity function is as follows:
in a specific embodiment, the parameter setting further needs to satisfy the following condition:
r≥2.5w
t＝round(std(I))
in one embodiment, after obtaining the wide lines, the wide lines need to be further refined, that is, the central lines of the wide lines are extracted as the image edges of the SAR image.
It should be understood that although the various steps in the flowcharts of fig. 1 and 2 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1 and 2 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 4, there is provided a SAR image edge detection apparatus combining wide line extraction, including: an intensity map construction module 402, a matrix acquisition module 404, an intensity enhancement module 406, a wide line extraction module 408, and an edge detection module 410, wherein:
an intensity map construction module 402, configured to obtain edge intensity information of the SAR image in each direction, and construct a preliminary edge intensity map;
a matrix obtaining module 404, configured to perform convolution operation on the preliminary edge intensity map and a preset gaussian kernel function to obtain a black plug matrix corresponding to the preliminary edge intensity map;
the intensity enhancing module 406 is configured to obtain an enhanced edge intensity map of the SAR image according to the eigenvalue and the eigenvector of the blackplug matrix;
a wide line extraction module 408, configured to filter the enhanced edge intensity map by using a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity map;
and an edge detection module 410, configured to extract a center line of the wide line to obtain an image edge of the SAR image.
In one embodiment, the intensity map building module 402 is further configured to perform convolution filtering on each dimension of the SAR image according to a preset smoothing filter, and normalize the convolution filtering to obtain edge intensities of the SAR image in each direction; obtaining the final edge strength of each pixel point in the SAR image according to the edge strength of the SAR image in each direction; and constructing a preliminary edge intensity map according to the final edge intensity of each pixel point in the SAR image.
In one embodiment, the matrix obtaining module 404 is further configured to obtain a preset wide line model, and obtain a wide line structure in the preliminary edge intensity map according to the wide line model; performing convolution according to the wide line structure and a preset Gaussian kernel function to obtain each order partial derivative of each pixel point in the wide line structure; and constructing a black plug matrix corresponding to the preliminary edge intensity graph according to each order partial derivative of each pixel point.
In one embodiment, the strength enhancing module 406 is further configured to obtain a feature vector with a largest absolute value in the blackplug matrix as a preset direction vector; for each pixel point in the preliminary edge intensity graph, calculating the maximum absolute value of the second-order directional derivative of each pixel point in the direction of the directional vector; and obtaining an enhanced edge intensity map of the SAR image according to the maximum absolute value corresponding to each pixel point.
In one embodiment, the wide line extraction module 408 is further configured to obtain a preset isotropic nonlinear filter and a preset circular mask; classifying the pixel points to a gray similar weighting mask when the gray value of the mask center of the circular mask and the gray value of the pixel points in the enhanced edge intensity graph meet a preset similarity function by using the isotropic nonlinear filter; acquiring preset constant weight of the circular mask; comparing the mask center with other pixel points in the circular mask according to the constant weight to obtain a gray level similar weighted mask weight of the mask center; and carrying out inversion operation according to the gray level similarity weighting mask weight at the center of the mask to obtain a wide line in the enhanced edge intensity image.
In one embodiment, the wide line extraction module 408 is further configured to use the gray value of the mask center, the gray value of the pixel point, and a preset threshold value as parameters in a hyperbolic tangent function to construct the similarity function.
In one embodiment, the wide line extraction module 408 is further configured to construct a step function according to the gray value of the mask center and the gray value of the pixel point; and constructing the similarity function by taking the step function and the threshold value as parameters in the hyperbolic tangent function.
For specific limitations of the SAR image edge detection apparatus combining wide line extraction, see the above limitations on the SAR image edge detection method combining wide line extraction, which are not described herein again. All or part of each module in the SAR image edge detection device combined with the wide line extraction can be realized by software, hardware and combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a terminal, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a wide line extraction combined SAR image edge detection method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 5 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In an embodiment, a computer device is provided, comprising a memory storing a computer program and a processor implementing the steps of the method in the above embodiments when the processor executes the computer program.
In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method in the above-mentioned embodiments.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent application shall be subject to the appended claims.
Claims (10)
1. A method for detecting edges of an SAR image by combining wide line extraction comprises the following steps:
acquiring edge intensity information of the SAR image in each direction, and constructing a preliminary edge intensity map;
performing convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a blackplug matrix corresponding to the preliminary edge intensity graph;
obtaining an enhanced edge intensity graph of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and extracting the central line of the wide line to obtain the image edge of the SAR image.
2. The method according to claim 1, wherein the obtaining of the edge intensity information of the SAR image in each direction and the constructing of the preliminary edge intensity map comprise:
carrying out convolution filtering on each dimension of the SAR image according to a preset smoothing filter, and normalizing to obtain the edge strength of the SAR image in each direction;
obtaining the final edge strength of each pixel point in the SAR image according to the edge strength of the SAR image in each direction;
and constructing a preliminary edge intensity map according to the final edge intensity of each pixel point in the SAR image.
3. The method according to claim 1, wherein performing convolution operation on the preliminary edge intensity map and a preset gaussian kernel function to obtain a blackplug matrix corresponding to the preliminary edge intensity map comprises:
acquiring a preset wide line model, and acquiring a wide line structure in the preliminary edge intensity map according to the wide line model;
performing convolution according to the wide line structure and a preset Gaussian kernel function to obtain each order partial derivative of each pixel point in the wide line structure;
and constructing a black plug matrix corresponding to the preliminary edge intensity graph according to each order partial derivative of each pixel point.
4. The method of claim 3, wherein obtaining the enhanced edge strength map of the SAR image according to the eigenvalues and eigenvectors of the blackplug matrix comprises:
acquiring a characteristic vector with the maximum absolute value in the black plug matrix as a preset direction vector;
for each pixel point in the preliminary edge intensity graph, calculating the maximum absolute value of the second-order directional derivative of each pixel point in the direction of the directional vector;
and obtaining an enhanced edge intensity map of the SAR image according to the maximum absolute value corresponding to each pixel point.
5. The method according to claim 1, wherein the filtering the enhanced edge intensity map by using a preset isotropic non-linear filter to obtain a wide line in the enhanced edge intensity map comprises:
acquiring a preset isotropic nonlinear filter and a preset circular mask;
classifying the pixel points to a gray similar weighting mask when the gray value of the mask center of the circular mask and the gray value of the pixel points in the enhanced edge intensity graph meet a preset similarity function by using the isotropic nonlinear filter;
acquiring preset constant weight of the circular mask;
comparing the mask center with other pixel points in the circular mask according to the constant weight to obtain a gray level similar weighted mask weight of the mask center;
and carrying out inversion operation according to the gray level similarity weighting mask weight at the center of the mask to obtain a wide line in the enhanced edge intensity image.
6. The method of claim 5, wherein the step of setting the similarity function comprises:
and constructing the similarity function by taking the gray value of the center of the mask, the gray value of the pixel point and a preset threshold value as parameters in the hyperbolic tangent function.
7. The method of claim 6, wherein the step of setting the similarity function comprises:
constructing a step function according to the gray value of the center of the mask and the gray value of the pixel point;
and constructing the similarity function by taking the step function and the threshold value as parameters in the hyperbolic tangent function.
8. An SAR image edge detection device combining wide line extraction is characterized by comprising:
the intensity map building module is used for obtaining the edge intensity information of the SAR image in each direction and building a preliminary edge intensity map;
the matrix acquisition module is used for carrying out convolution operation on the preliminary edge intensity graph and a preset Gaussian kernel function to obtain a black plug matrix corresponding to the preliminary edge intensity graph;
the intensity enhancing module is used for obtaining an enhanced edge intensity image of the SAR image according to the eigenvalue and the eigenvector of the black plug matrix;
the wide line extraction module is used for filtering the enhanced edge intensity graph by adopting a preset isotropic nonlinear filter to obtain a wide line in the enhanced edge intensity graph;
and the edge detection module is used for extracting the central line of the wide line to obtain the image edge of the SAR image.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910976380.5A CN110796674A (en) | 2019-10-15 | 2019-10-15 | SAR image edge detection method and device combining wide line extraction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910976380.5A CN110796674A (en) | 2019-10-15 | 2019-10-15 | SAR image edge detection method and device combining wide line extraction |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110796674A true CN110796674A (en) | 2020-02-14 |
Family
ID=69440363
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910976380.5A Pending CN110796674A (en) | 2019-10-15 | 2019-10-15 | SAR image edge detection method and device combining wide line extraction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110796674A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106650697A (en) * | 2016-12-30 | 2017-05-10 | 亿嘉和科技股份有限公司 | Instrument scale recognition method |
| CN107430762A (en) * | 2015-02-25 | 2017-12-01 | 高通股份有限公司 | Digital zooming method and system |
| CN109712156A (en) * | 2018-12-12 | 2019-05-03 | 西安电子科技大学 | A kind of SAR image edge detection method of low error rate |
-
2019
- 2019-10-15 CN CN201910976380.5A patent/CN110796674A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107430762A (en) * | 2015-02-25 | 2017-12-01 | 高通股份有限公司 | Digital zooming method and system |
| CN106650697A (en) * | 2016-12-30 | 2017-05-10 | 亿嘉和科技股份有限公司 | Instrument scale recognition method |
| CN109712156A (en) * | 2018-12-12 | 2019-05-03 | 西安电子科技大学 | A kind of SAR image edge detection method of low error rate |
Non-Patent Citations (2)
| Title |
|---|
| LAURA LIU ET AL.: ""Detecting Wide Lines Using Isotropic Nonlinear Filtering"", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
| YUXIAO LUO ET AL.: ""Improved ROEWA SAR Image Edge Detector Based on Curvilinear Structures Extraction"", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111860670B (en) | Domain adaptive model training method, image detection method, device, equipment and medium | |
| Liu et al. | Additive white Gaussian noise level estimation in SVD domain for images | |
| CN111080628A (en) | Image tampering detection method and device, computer equipment and storage medium | |
| US9489722B2 (en) | Method and apparatus for implementing image denoising | |
| Lo et al. | Edge-preserving depth map upsampling by joint trilateral filter | |
| Ray | Unsupervised edge detection and noise detection from a single image | |
| CN109903272B (en) | Target detection method, device, equipment, computer equipment and storage medium | |
| CN114155365B (en) | Model training method, image processing method and related device | |
| JP2015099559A (en) | Image processing apparatus, image processing method, and program | |
| CN113781406B (en) | Scratch detection method and device for electronic component and computer equipment | |
| Kumar et al. | A conventional study of edge detection technique in digital image processing | |
| CN110533632B (en) | Image blurring tampering detection method and device, computer equipment and storage medium | |
| CN112990313A (en) | Hyperspectral image anomaly detection method and device, computer equipment and storage medium | |
| CN112381727A (en) | Image denoising method and device, computer equipment and storage medium | |
| CN110956602B (en) | Method and device for determining change area and storage medium | |
| CN111598088B (en) | Target detection method, device, computer equipment and readable storage medium | |
| CN116485834A (en) | Method, device, equipment and medium for detecting infrared dim target | |
| Krylov et al. | False discovery rate approach to unsupervised image change detection | |
| CN111539969B (en) | Image edge detection method, device, computer equipment and storage medium | |
| CN110796674A (en) | SAR image edge detection method and device combining wide line extraction | |
| CN115035350B (en) | Edge detection enhancement-based method for detecting small objects on air-ground and ground background | |
| Tsai et al. | Image co-saliency detection via locally adaptive saliency map fusion | |
| Hao et al. | Active cues collection and integration for building extraction with high-resolution color remote sensing imagery | |
| Islam et al. | A study on edge detection techniques for natural image segmentation | |
| CN114359232A (en) | Image change detection method and device based on context covariance matrix |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200214 |