CN113538240A - SAR image superpixel generation method and device, computer equipment and storage medium - Google Patents

SAR image superpixel generation method and device, computer equipment and storage medium Download PDF

Info

Publication number
CN113538240A
CN113538240A CN202110807939.9A CN202110807939A CN113538240A CN 113538240 A CN113538240 A CN 113538240A CN 202110807939 A CN202110807939 A CN 202110807939A CN 113538240 A CN113538240 A CN 113538240A
Authority
CN
China
Prior art keywords
pixels
sar image
pixel
seed
initial
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.)
Granted
Application number
CN202110807939.9A
Other languages
Chinese (zh)
Other versions
CN113538240B (en
Inventor
张亮
刘涛
栗毅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN202110807939.9A priority Critical patent/CN113538240B/en
Publication of CN113538240A publication Critical patent/CN113538240A/en
Application granted granted Critical
Publication of CN113538240B publication Critical patent/CN113538240B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformation in the plane of the image
    • G06T3/40—Scaling the whole image or part thereof
    • G06T3/4053—Super resolution, i.e. output image resolution higher than sensor resolution
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • 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

Abstract

The application relates to a method and a device for generating super pixels of an SAR image, computer equipment and a storage medium. The method comprises the following steps: the method comprises the steps of selecting a plurality of unmarked pixels in an SAR image as seeds, calculating the difference between the unmarked pixels and various sub-adjacent pixels, marking the pixels with the difference smaller than a threshold value with the same marks as the seeds, clustering the whole SAR image by the method to obtain initial superpixels with different labels, merging the pixels with the number smaller than the threshold value in the initial superpixels, and taking the merged initial superpixels as final superpixels of the SAR image. The method has the characteristics of self-adaptability and quickness, and can effectively provide support for the subsequent SAR image processing task based on the superpixel.

Description

SAR image superpixel generation method and device, computer equipment and storage medium
Technical Field
The present application relates to the field of synthetic aperture radar image processing technologies, and in particular, to a method and an apparatus for generating a super-pixel of an SAR image, a computer device, and a storage medium.
Background
Synthetic Aperture Radar (SAR) remote sensing images become an important component of current remote sensing data with all-weather, all-day, high resolution and large-area data acquisition capability, and have been widely applied in the aspects of resources, environment, urban construction, military and the like.
In recent years, superpixel-based approaches have received increasing researchers' attention in Synthetic Aperture Radar (SAR) image interpretation because superpixels capture local image information better than pixels and reduce the computational complexity of subsequent SAR image processing tasks (e.g., target detection, image segmentation, image classification, etc.).
Disclosure of Invention
In view of the above, it is necessary to provide a method, an apparatus, a computer device, and a storage medium for generating superpixels of a SAR image, which can generate superpixels quickly.
A SAR image superpixel generation method, the method comprising:
acquiring an SAR image;
selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same marks to the pixels and the seeds when the difference is smaller than a first threshold value until a termination condition is met, finishing the marking, and taking the seeds and the pixels with the same marks as the seeds as initial superpixels with the same category;
selecting another unmarked pixel in the SAR image as a seed, giving a new mark to the seed, marking the pixel belonging to the same category as the seed to obtain an initial superpixel of another category until all the pixels in the SAR image are marked to obtain a plurality of initial superpixels with different categories;
and combining the initial superpixels with the number of pixels smaller than a second threshold value in each initial superpixel with the adjacent initial superpixels with the minimum difference to obtain the final superpixel of the SAR image.
In one embodiment, a plurality of pixels are selected in advance from the SAR image as a seed candidate set, and pixels in the seed candidate set are selected as seeds.
In one embodiment, the difference between the seed and the unlabeled pixel is calculated using the following formula:
Ω(i,j)＝σ(i,j)+|H(i)-H(j)|·σ(k,j)
Figure BDA0003167105090000021
where Ω (i, j) represents the difference between the seed pixel i and the unmarked pixel j, and the subscript PiAnd subscript PjRepresenting two regions, I, centered on a seed pixel I and an unmarked pixel jPiAnd IPjRepresents PiAnd PjAverage intensity of two regions, M is the number of pixels in a region, L is the view of the SAR image, ξ (i) and ξ (j) represent the edge intensities of the seed pixel i and the unmarked pixel j, h (i) and h (j) represent the homogeneity of the seed pixel i and the unmarked pixel j, k represents the central pixel in the set of pixels with the same mark as the seed pixel i;
homogeneity of the seed pixel i and the unmarked pixel j is determined by the PiAnd PjThe coefficient of variation of (a) is calculated.
In one embodiment, the disparity between two initial superpixels is calculated using the following formula:
Figure BDA0003167105090000022
Figure BDA0003167105090000023
wherein omega (SP)m,SPn) Representing an initial superpixel SPmAnd an initial superpixel SPnThe difference value between, size (·) denotes the size of the initial superpixel, L is the view of the SAR image,
Figure BDA0003167105090000031
and
Figure BDA0003167105090000032
represents SPmAnd SPnAverage intensity of two pixels, ξ (SP)m) And xi (SP)n) Represents SPmAnd SPnAverage edge strength of H (SP)m) And H (SP)n) Represents SPmAnd SPnHomogeneity of (a);
the SPmAnd SPnHomogeneity is determined by the SPmAnd SPnThe coefficient of variation of (a) is calculated.
In one embodiment, after the SAR image is acquired, the SAR image is further processed to obtain an edge intensity map corresponding to the SAR image;
the edge intensities of the seed pixel i and the unmarked pixel j, and the initial superpixel SPmAnd an initial superpixel SPnThe average edge intensity of the SAR image is obtained by an edge intensity map corresponding to the SAR image.
In one embodiment, the termination condition includes:
pixels not newly marked, or
Four neighboring pixels of the newly marked pixel have all been marked, or
The number of the pixels marked at this time is greater than or equal to the ratio of the total number of the pixels of the SAR image to the preset total number of the super pixels.
In one embodiment, the second threshold is related to a total number of pixels of the SAR image and a preset total number of superpixels.
A SAR image superpixel generation apparatus, the apparatus comprising:
the SAR image acquisition module is used for acquiring an SAR image;
the adjacent pixel marking module is used for selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in the four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
the initial super-pixel obtaining module is used for calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same mark to the pixels and the seeds when the difference is smaller than a first threshold value until the termination condition is met, finishing the marking, and taking the seeds and the pixels with the same mark as the seeds as the initial super-pixels with the same category;
a plurality of initial super-pixel obtaining modules, configured to select another unmarked pixel in the SAR image as a seed, give the seed a new mark, mark a pixel belonging to the same category as the seed, and obtain another category of initial super-pixels until all pixels in the SAR image are marked, so as to obtain a plurality of initial super-pixels with different categories;
and a final superpixel obtaining module, configured to combine the initial superpixel with the number of pixels smaller than the second threshold in each initial superpixel with the adjacent initial superpixel with the smallest difference, and obtain a final superpixel of the SAR image.
A computer device comprising a memory storing a computer program and a processor implementing the steps of the SAR image superpixel generation method described above when the computer program is executed.
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 SAR image superpixel generation method described above.
According to the SAR image superpixel generation method, the device, the computer equipment and the storage medium, a plurality of unmarked pixels are selected from the SAR image as seeds, the difference between the pixels adjacent to various seeds is calculated, the pixels with the difference smaller than the threshold are marked with the same marks as the seeds, the whole SAR image is clustered by the method, initial superpixels with different labels are obtained, the pixels with the number smaller than the threshold in the initial superpixels are merged, and the initial superpixels obtained after merging are used as final superpixels of the SAR image. The SAR image superpixel generation method has the characteristics of sub-adaptability and high speed, and can effectively provide support for the subsequent SAR image processing task based on the superpixel.
Drawings
FIG. 1 is a schematic flow chart of a method for generating superpixels of an SAR image in one embodiment;
FIG. 2 is a flowchart illustrating the actual operation steps of the SAR image super-pixel generation method in one embodiment;
FIG. 3 is a block diagram of a SAR image super-pixel generation device in one embodiment;
FIG. 4 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.
As shown in fig. 1, there is provided a method for generating superpixels of an SAR image, comprising the following steps:
s100, acquiring an SAR image;
step S110, selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is less than a first threshold value;
step S120, calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the pixels and the seeds the same mark when the difference is smaller than a first threshold value until the termination condition is met, finishing the marking, and taking the seeds and the pixels with the same mark as the seeds as initial superpixels with the same category;
step S130, selecting another unmarked pixel in the SAR image as a seed, giving a new mark to the seed, marking the pixel belonging to the same category as the seed to obtain another category of initial superpixels until all the pixels in the SAR image are marked to obtain a plurality of initial superpixels with different categories;
step S140, merging the initial superpixels with the number of pixels smaller than the second threshold value in each initial superpixel with the adjacent initial superpixels with the minimum difference, and obtaining the final superpixel of the SAR image.
The SAR image superpixel generation can be understood as that a plurality of areas are formed by a series of pixels which are adjacent in position and have similar characteristics of color, brightness, texture and the like in the SAR image, and the areas comprise a plurality of pixels, so that each area is a superpixel. The SAR image superpixel can better capture local image information than pixels, thereby reducing subsequent SAR image processing tasks such as target detection, image segmentation, image classification, and the like.
In the method, all pixels are clustered firstly based on a clustering method to obtain a plurality of initial superpixels, in order to ensure that the obtained superpixels are more regular and more compact in edge fitting, the number of pixels in each superpixel is as large as possible so as to improve the processing speed of subsequent application, and then the pixels with less data are combined with the adjacent initial superpixels to obtain the final superpixels.
In steps S110 to S120, a process of marking a plurality of pixels related to a seed is performed, and after the present marking is finished, an initial superpixel is obtained.
And when the difference is smaller than a preset threshold value, the pixel and the seed are proved to belong to the same target of the SAR image with high probability, and the pixel is marked as the seed. And then, for the newly marked pixels, the number of the newly marked pixels may be one or more, the difference between the unmarked pixels and the seeds in the four adjacent pixels is the same, and the pixels with the difference smaller than the threshold value are marked with the same marks as the seeds, so that the operation steps are repeated until the termination condition is met, the marking is completed, namely the pixels with the same type as the seeds are found out and the same marks are given, and the clustering of the initial superpixels is completed.
In step S130, an unmarked pixel is selected from the SAR image as a seed, and a new mark is given, and after steps S110 to S120 are repeated and the marking is completed, the same pixels that are the same type as the seed are found and the same mark is given, thereby completing the clustering of the second initial superpixel. Steps S110 to S120 are repeated in this way until all pixels in the SAR image are marked, and a plurality of pixel regions given different marks, that is, a plurality of initial superpixels, can be obtained. Different labels mean that the targets are different in the respective categories, i.e. in the respective initial superpixels.
In this embodiment, the termination conditions include: pixels which are not marked newly, or four adjacent pixels of the newly marked pixels are marked, or the number of the pixels marked this time is larger than or equal to the ratio of the total number of the pixels of the SAR image to the total number of the preset superpixels.
In this embodiment, a plurality of pixels in the SAR image are pre-selected as a seed candidate set, and pixels in the seed candidate set are selected as seeds. The number of seeds in the seed candidate set can be determined according to the size definition of the SAR image, and each seed can be manually specified on the SAR image by people.
In steps S110 and S120, the following formula is used to calculate the difference between the seed and the unmarked pixel:
Ω(i,j)＝σ(i,j)+|H(i)-H(j)|·σ(k,j)(1)
Figure BDA0003167105090000071
in equation (1), Ω (i, j) represents the difference value between the seed pixel i and the unmarked pixel j, subscriptPiAnd subscript PjRepresenting two regions centered on a seed pixel i and an unmarked pixel j,
Figure BDA0003167105090000072
and
Figure BDA0003167105090000073
represents PiAnd PjThe average intensity of the two regions, M is the number of pixels in the region, L is the view of the SAR image, ξ (i) and ξ (j) represent the edge intensities of the seed pixel i and unmarked pixel j, h (i) and h (j) represent the homogeneity of the seed pixel i and unmarked pixel j, and k represents the center pixel in the set of pixels with the same label as the seed pixel i. Homogeneity of the seed pixel i and the unmarked pixel j is determined by PiAnd PjThe coefficient of variation of (a) is calculated.
In one implementation, the size of the two regions centered on the seed pixel i and the unmarked pixel j may be selected to be 5 × 5 size.
In step S140, the plurality of initial superpixels are further merged, and the initial superpixels with the number of pixels less than the second threshold value are merged into other initial superpixels. And during merging, calculating the initial superpixel which is adjacent to the initial superpixel and has the minimum difference, merging, so as to further reduce the number of the initial superpixels and expand the number of the pixels, wherein the merged initial superpixel is used as the final superpixel of the SAR image.
In this embodiment, the following formula is used to calculate the difference between two adjacent initial superpixels:
Figure BDA0003167105090000074
in the formula (2), Ω (SP)m,SPn) Representing an initial superpixel SPmAnd an initial superpixel SPnThe difference value between, size (·) denotes the size of the initial superpixel, L is the view of the SAR image,
Figure BDA0003167105090000075
and
Figure BDA0003167105090000076
represents SPmAnd SPnAverage intensity of two pixels, ξ (SP)m) And xi (SP)n) Represents SPmAnd SPnAverage edge strength of H (SP)m) And H (SP)n) Represents SPmAnd SPnHomogeneity of the composition. SPmAnd SPnHomogeneity is determined by the SPmAnd SPnThe coefficient of variation of (a) is calculated.
In the present embodiment, the second threshold is related to the total number of pixels of the SAR image and the preset total number of superpixels. For example, the second threshold may be a ratio S/K of the total number of pixels S of the SAR image and a preset total number of superpixels K, or 0.2S/K.
In step S100, after the SAR image is acquired, the SAR image is further processed to obtain an edge intensity map corresponding to the SAR image. Thus in equations (1) and (2), the edge intensities of the seed pixel i and unmarked pixel j are calculated, as well as the initial superpixel SPmAnd an initial superpixel SPnThe average edge intensities of the SAR images are obtained by the edge intensity image corresponding to the SAR images.
The application also provides a step in actual operation aiming at the SAR image superpixel generation method, as shown in fig. 2, the specific implementation steps are as follows:
step S1) the SAR image is processed using an edge detector with a gaussian-shaped window, resulting in an edge intensity map.
Step S2), clustering all pixels of the SAR image by applying a similar DBSCAN algorithm, and the specific process is as follows:
step S2-1) all pixels of the SAR image are regarded as not marked, and a certain number of preset pixels are selected as a seed candidate set.
Step S2-2) selects an unmarked pixel as a seed (preferentially selected from the seed candidate set), and gives a new mark, and initializes an empty set C.
Step S2-3) puts the unmarked pixels of the four neighboring pixels of the seed into the set C.
Step S2-4), when the set C is empty or the number of marked pixels in the mark of the current round is greater than or equal to S/K, the mark of the current round is stopped, and the step S2-2) is skipped to start a new round of marking. Wherein S is the total number of pixels of the SAR image, and K is the preset total number of super pixels.
Step S2-5) extracts a pixel P from the set C (when extracting, the element is deleted from the set C), and calculates the difference between the seed and the pixel P, wherein the calculation formula of the difference adopts the above formula (1). When the difference is smaller than the threshold value, giving the pixel P the same mark as the seed, putting the unmarked pixels of the four adjacent pixels into the set C, and jumping to the step S2-4); when the difference is larger than the threshold, go directly to step S2-4).
Step S2-6) repeats steps S2-2) through S2-5) until all pixels are marked. Different labels are of different classes.
Step S3) the superpixels obtained after the clustering in the step S2) are used as initial superpixels, and the initial superpixels are merged to obtain final superpixels. The specific process is as follows:
step S3-1) puts the initial superpixel into the set to be merged D when the number of pixels in the superpixel is less than a given threshold (0.2 × S/K).
Step S3-2) extracts one super pixel SP from the set D (the element is deleted from the set D when the super pixel SP is extracted), and calculates the difference between the SP and the adjacent super pixels, wherein the calculation formula of the difference is the above formula (2). And selecting the adjacent superpixel with the minimum difference to be combined with the SP to form a new superpixel.
Step S3-3) repeats steps S3-1) to S2-2) until the set D is empty.
According to the SAR image superpixel generation method, a plurality of initial superpixels are obtained from the SAR image through a clustering method, and then the initial superpixels are combined to generate the final superpixel of the SAR image.
It should be understood that, although the steps in the flowchart of fig. 1 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 a portion of the steps in fig. 1 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 a portion of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 3, there is provided a SAR image superpixel generation apparatus, including: an SAR image acquisition module 200, an adjacent pixel labeling module 210, an initial superpixel obtaining module 220, a plurality of initial superpixel obtaining modules 230, and a final superpixel obtaining module 240, wherein:
an SAR image acquisition module 200 for acquiring an SAR image;
a neighboring pixel labeling module 210, configured to select an unlabeled pixel in the SAR image as a seed, give the seed a new label, calculate a difference between the seed and an unlabeled pixel in four neighboring pixels of the seed, and give the pixel a label that is the same as the seed when the difference is smaller than a first threshold;
an initial super-pixel obtaining module 220, configured to calculate differences between unmarked pixels and seeds in four adjacent pixels of a newly marked pixel, and when the differences are smaller than a first threshold, give the same mark to the pixels and the seeds until a termination condition is met, complete the marking, and use the seeds and the pixels with the same mark as the seeds as initial super-pixels with the same category;
a plurality of initial super-pixel obtaining modules 230, configured to select another unmarked pixel in the SAR image as a seed, give a new mark to the seed, mark a pixel belonging to the same category as the seed, and obtain another category of initial super-pixels until all pixels in the SAR image are marked, so as to obtain a plurality of initial super-pixels with different categories;
a final super-pixel obtaining module 240, configured to combine the initial super-pixels, of which the number of pixels is smaller than the second threshold, of each initial super-pixel with the adjacent initial super-pixel with the minimum difference, so as to obtain a final super-pixel of the SAR image.
For specific limitations of the SAR image superpixel generation device, reference may be made to the above limitations of the SAR image superpixel generation method, and details are not repeated here. The modules in the SAR image superpixel generation device can be wholly or partially realized by software, hardware and a 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, and its internal structure diagram may be as shown in fig. 4. 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 SAR image superpixel generation 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. 4 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 one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring an SAR image;
selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same marks to the pixels and the seeds when the difference is smaller than a first threshold value until a termination condition is met, finishing the marking, and taking the seeds and the pixels with the same marks as the seeds as initial superpixels with the same category;
selecting another unmarked pixel in the SAR image as a seed, giving a new mark to the seed, marking the pixel belonging to the same category as the seed to obtain an initial superpixel of another category until all the pixels in the SAR image are marked to obtain a plurality of initial superpixels with different categories;
and combining the initial superpixels with the number of pixels smaller than a second threshold value in each initial superpixel with the adjacent initial superpixels with the minimum difference to obtain the final superpixel of the SAR image.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring an SAR image;
selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same marks to the pixels and the seeds when the difference is smaller than a first threshold value until a termination condition is met, finishing the marking, and taking the seeds and the pixels with the same marks as the seeds as initial superpixels with the same category;
selecting another unmarked pixel in the SAR image as a seed, giving a new mark to the seed, marking the pixel belonging to the same category as the seed to obtain an initial superpixel of another category until all the pixels in the SAR image are marked to obtain a plurality of initial superpixels with different categories;
and combining the initial superpixels with the number of pixels smaller than a second threshold value in each initial superpixel with the adjacent initial superpixels with the minimum difference to obtain the final superpixel of the SAR image.
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 shall be subject to the appended claims.

Claims (10)

  1. A SAR image superpixel generation method, characterized in that the method comprises:
    acquiring an SAR image;
    selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
    calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same marks to the pixels and the seeds when the difference is smaller than a first threshold value until a termination condition is met, finishing the marking, and taking the seeds and the pixels with the same marks as the seeds as initial superpixels with the same category;
    selecting another unmarked pixel in the SAR image as a seed, giving a new mark to the seed, marking the pixel belonging to the same category as the seed to obtain an initial superpixel of another category until all the pixels in the SAR image are marked to obtain a plurality of initial superpixels with different categories;
    and combining the initial superpixels with the number of pixels smaller than a second threshold value in each initial superpixel with the adjacent initial superpixels with the minimum difference to obtain the final superpixel of the SAR image.
  2. 2. The SAR image superpixel generation method of claim 1, characterized in that, a plurality of pixels are pre-selected in the SAR image as a seed candidate set, and pixels in the seed candidate set are selected as seeds.
  3. 3. The SAR image superpixel generation method according to claim 1, characterized in that the dissimilarity between seeds and unlabeled pixels is calculated using the following formula:
    Ω(i,j)＝σ(i,j)+|H(i)-H(j)|·σ(k,j)
    Figure FDA0003167105080000011
    where Ω (i, j) represents the difference between the seed pixel i and the unmarked pixel j, and the subscript PiAnd subscript PjRepresenting two regions centered on a seed pixel i and an unmarked pixel j,
    Figure FDA0003167105080000012
    and
    Figure FDA0003167105080000013
    represents PiAnd PjAverage intensity of two regions, M is the number of pixels in a region, L is the view of the SAR image, and xi (i) and xi (j) represent seed pixels i and xi (j)Edge intensity of unlabeled pixel j, h (i) and h (j) represent homogeneity of seed pixel i and unlabeled pixel j, k represents a center pixel in a set of pixels with the same label as seed pixel i;
    homogeneity of the seed pixel i and the unmarked pixel j is determined by the PiAnd PjThe coefficient of variation of (a) is calculated.
  4. 4. The SAR image superpixel generation method according to claim 1, characterized in that the dissimilarity between two initial superpixels is calculated using the following formula:
    Figure FDA0003167105080000021
    wherein omega (SP)m,SPn) Representing an initial superpixel SPmAnd an initial superpixel SPnThe difference value between, size (·) denotes the size of the initial superpixel, L is the view of the SAR image,
    Figure FDA0003167105080000022
    and
    Figure FDA0003167105080000023
    represents SPmAnd SPnAverage intensity of two pixels, ξ (SP)m) And xi (SP)n) Represents SPmAnd SPnAverage edge strength of H (SP)m) And H (SP)n) Represents SPmAnd SPnHomogeneity of (a);
    the SPmAnd SPnHomogeneity is determined by the SPmAnd SPnThe coefficient of variation of (a) is calculated.
  5. 5. The SAR image superpixel generation method according to claim 3 or 4,
    after the SAR image is obtained, processing the SAR image to obtain an edge intensity map corresponding to the SAR image;
    the edge intensities of the seed pixel i and the unmarked pixel j, and the initial superpixel SPmAnd an initial superpixel SPnThe average edge intensity of the SAR image is obtained by an edge intensity map corresponding to the SAR image.
  6. 6. The SAR image superpixel generation method according to claim 1, characterized in that said termination condition comprises:
    pixels not newly marked, or
    Four neighboring pixels of the newly marked pixel have all been marked, or
    The number of the pixels marked at this time is greater than or equal to the ratio of the total number of the pixels of the SAR image to the preset total number of the super pixels.
  7. 7. The SAR image superpixel generation method according to claim 1, characterized in that said second threshold value is related to the total number of pixels of said SAR image and a preset total number of superpixels.
  8. 8. An apparatus for generating superpixels for a SAR image, the apparatus comprising:
    the SAR image acquisition module is used for acquiring an SAR image;
    the adjacent pixel marking module is used for selecting an unmarked pixel in the SAR image as a seed, giving a new mark to the seed, calculating the difference between the seed and the unmarked pixels in the four adjacent pixels, and giving the same mark to the pixel and the seed when the difference is smaller than a first threshold value;
    the initial super-pixel obtaining module is used for calculating the difference between the unmarked pixels and the seeds in the four adjacent pixels of the newly marked pixels, giving the same mark to the pixels and the seeds when the difference is smaller than a first threshold value until the termination condition is met, finishing the marking, and taking the seeds and the pixels with the same mark as the seeds as the initial super-pixels with the same category;
    a plurality of initial super-pixel obtaining modules, configured to select another unmarked pixel in the SAR image as a seed, give the seed a new mark, mark a pixel belonging to the same category as the seed, and obtain another category of initial super-pixels until all pixels in the SAR image are marked, so as to obtain a plurality of initial super-pixels with different categories;
    and a final superpixel obtaining module, configured to combine the initial superpixel with the number of pixels smaller than the second threshold in each initial superpixel with the adjacent initial superpixel with the smallest difference, and obtain a final superpixel of the SAR image.
  9. 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. 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.
CN202110807939.9A 2021-07-16 2021-07-16 SAR image superpixel generation method and device, computer equipment and storage medium Active CN113538240B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110807939.9A CN113538240B (en) 2021-07-16 2021-07-16 SAR image superpixel generation method and device, computer equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110807939.9A CN113538240B (en) 2021-07-16 2021-07-16 SAR image superpixel generation method and device, computer equipment and storage medium

Publications (2)

Publication Number Publication Date
CN113538240A true CN113538240A (en) 2021-10-22
CN113538240B CN113538240B (en) 2022-09-23

Family

ID=78099876

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110807939.9A Active CN113538240B (en) 2021-07-16 2021-07-16 SAR image superpixel generation method and device, computer equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113538240B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180012365A1 (en) * 2015-03-20 2018-01-11 Ventana Medical Systems, Inc. System and method for image segmentation
CN109389601A (en) * 2018-10-19 2019-02-26 山东大学 Color image superpixel segmentation method based on similitude between pixel
CN109712153A (en) * 2018-12-25 2019-05-03 杭州世平信息科技有限公司 A kind of remote sensing images city superpixel segmentation method
CN111008981A (en) * 2019-12-26 2020-04-14 中国人民解放军国防科技大学 Method, system, device and computer readable medium for segmenting polarimetric synthetic aperture radar image
CN111260596A (en) * 2020-01-09 2020-06-09 山东财经大学 Anti-noise rapid image super-pixel automatic generation method, device and readable storage medium
CN112529910A (en) * 2020-12-08 2021-03-19 电科云(北京)科技有限公司 SAR image rapid superpixel merging and image segmentation method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180012365A1 (en) * 2015-03-20 2018-01-11 Ventana Medical Systems, Inc. System and method for image segmentation
CN109389601A (en) * 2018-10-19 2019-02-26 山东大学 Color image superpixel segmentation method based on similitude between pixel
CN109712153A (en) * 2018-12-25 2019-05-03 杭州世平信息科技有限公司 A kind of remote sensing images city superpixel segmentation method
CN111008981A (en) * 2019-12-26 2020-04-14 中国人民解放军国防科技大学 Method, system, device and computer readable medium for segmenting polarimetric synthetic aperture radar image
CN111260596A (en) * 2020-01-09 2020-06-09 山东财经大学 Anti-noise rapid image super-pixel automatic generation method, device and readable storage medium
CN112529910A (en) * 2020-12-08 2021-03-19 电科云(北京)科技有限公司 SAR image rapid superpixel merging and image segmentation method

Non-Patent Citations (7)

* Cited by examiner, † Cited by third party
Title
D. XIANG, T. TANG, L. ZHAO AND Y. SU: "Superpixel Generating Algorithm Based on Pixel Intensity and Location Similarity forSAR Image Classification", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 *
D. XIANG, T. TANG, S. QUAN, D. GUAN AND Y. SU: "Adaptive Superpixel Generation for SAR Images With Linear Feature Clustering andEdge Constraint", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》 *
H. HU, B. LIU, W. GUO, Z. ZHANG AND W. YU: "Superpixel generation for SAR images based on DBSCAN clustering and probabilistic patch-based similarity", 《2017 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS)》 *
冯于珍等: "融合边缘信息的合成孔径雷达图像超像素分割算法", 《科学技术与工程》 *
崔兴超; 粟毅; 陈思伟: "融合极化旋转域特征和超像素技术的极化SAR舰船检测", 《雷达学报》 *
张亮; 李禹; 计科峰; 粟毅: "SAR 图像局部自适应ACCA-CFAR检测算法", 《电子与信息学报》 *
蒋婷等: "极化SAR图像分割方法研究", 《无线互联科技》 *

Also Published As

Publication number Publication date
CN113538240B (en) 2022-09-23

Similar Documents

Publication Publication Date Title
CN109903272B (en) Target detection method, device, equipment, computer equipment and storage medium
CN110490902B (en) Target tracking method and device applied to smart city and computer equipment
CN108986152B (en) Foreign matter detection method and device based on difference image
CN113496150B (en) Dense target detection method and device, storage medium and computer equipment
CN114092833B (en) Remote sensing image classification method and device, computer equipment and storage medium
CN111047088A (en) Prediction image acquisition method and device, computer equipment and storage medium
CN108470185A (en) The atural object annotation equipment and method of satellite image
CN111275743A (en) Target tracking method, device, computer readable storage medium and computer equipment
CN112381107A (en) Article X-ray detection method and device based on deep learning and computer equipment
CN112241646A (en) Lane line recognition method and device, computer equipment and storage medium
CN113989656A (en) Event interpretation method and device for remote sensing video, computer equipment and storage medium
CN110309855B (en) Training method for image segmentation, computer device and storage medium
CN115526846A (en) Joint detection-based crack detection method and device and computer equipment
CN104732534A (en) Method and system for matting conspicuous object in image
CN113537020A (en) Complex SAR image target identification method based on improved neural network
CN110163864B (en) Image segmentation method and device, computer equipment and storage medium
CN113538240B (en) SAR image superpixel generation method and device, computer equipment and storage medium
CN110889432B (en) Feature point matching method and device, computer equipment and storage medium
CN110516559B (en) Target tracking method and device suitable for accurate monitoring and computer equipment
CN113449586A (en) Target detection method, target detection device, computer equipment and storage medium
CN112418033A (en) Landslide slope surface segmentation and identification method based on mask rcnn neural network
CN116310832A (en) Remote sensing image processing method, device, equipment, medium and product
CN107292840B (en) Image restoration method and device, computer-readable storage medium and terminal
CN116310899A (en) YOLOv 5-based improved target detection method and device and training method
CN114372936A (en) Infrared dim target detection method based on image inpainting technology

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CB03 Change of inventor or designer information

Inventor after: Zhang Liang

Inventor after: Liu Tao

Inventor after: Su Yi

Inventor before: Zhang Liang

Inventor before: Liu Tao

Inventor before: Li Yi

CB03 Change of inventor or designer information