CN107862271B - Detection method of ship target - Google Patents
Detection method of ship target Download PDFInfo
- Publication number
- CN107862271B CN107862271B CN201711049084.8A CN201711049084A CN107862271B CN 107862271 B CN107862271 B CN 107862271B CN 201711049084 A CN201711049084 A CN 201711049084A CN 107862271 B CN107862271 B CN 107862271B
- Authority
- CN
- China
- Prior art keywords
- matrix
- ship
- strong scattering
- sar image
- window
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Remote Sensing (AREA)
- Astronomy & Astrophysics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Radar Systems Or Details Thereof (AREA)
- Image Processing (AREA)
Abstract
The invention belongs to the technical field of radar signal processing, and discloses a method for detecting a ship target. However, in the actual processing, the port target on land and the strong scattering points in the urban area near the coast affect the detection of the ship target, so before the processing, the template matching method is used to completely separate the sea and land area, and then the ship target is effectively identified by setting the threshold for the pure sea surface and the ship.
Description
Technical Field
The invention belongs to the technical field of radar signal processing, and particularly relates to a method for detecting a ship target.
Background
The space-borne SAR is an effective means for observing the earth from space, has the advantages of all-time, all-weather, multi-band and high resolution, can accurately survey and draw the terrain and the landform in detail, acquires the information of the earth surface, and finally generates a high-resolution map of a target scene.
Seventy percent of the earth's area is covered by oceans, and the vast expanse of sea makes satellite-borne SAR a powerful tool for mapping earth. As an active remote sensing technology with a radiation source, the SAR image supplements the optical image with the unique advantages thereof, the defects of the optical image are perfectly compensated under the severe weather conditions of night, cloud, rain, fog and the like, and the earth can be effectively observed.
The commonly used method for detecting the ship target by the satellite-borne SAR image at present comprises the following steps: a detection method based on edge information, a detection method based on prior information of a port and the like. For marine ship targets in remote sensing images, the background of the marine ship targets is basically a natural scene mainly comprising sea surface, and the marine ship targets are obviously different from the marine ship targets in characteristics such as gray scale, texture and the like, so that the method based on edge detection obtains a better segmentation effect. However, for ships parked near ports, the background is no longer a single sea surface scene, and a large number of artificial targets, especially docks, have small difference from the gray level characteristics of the ships, and the dock and the ship targets cannot be distinguished by the edge information detection method, so that the false alarm probability is increased. The port contour is pre-stored in a template mode based on the prior information of the port, the port area is separated after geographic coordinate matching, and then subsequent target detection is carried out under the sea surface background. However, if the accuracy of the geographical coordinates of the port is too low or the port layout becomes small, which will result in a local misalignment of the template, the probability of detection of the ship target will be reduced.
Disclosure of Invention
In view of the defects of the prior art, the invention aims to provide a method for detecting a ship target, which can be used for efficiently detecting the ship target on the sea surface.
And radar echo data downloaded by the satellite-borne SAR are processed by an imaging algorithm to obtain an SAR image. Different from an optical image, the SAR image is convenient for human eye identification, and the backscattering intensity of a target in a scene reflected by each pixel point in the SAR image is a gray image displayed according to the amplitude.
According to the characteristic of the great difference of the ship target and the sea surface in the backscattering intensity, the ship on the sea surface can be efficiently identified by utilizing the difference characteristic of the amplitude of the ship target and the amplitude of the sea surface in the SAR image. However, in the actual processing, the port target on land and the strong scattering points in the urban area near the coast affect the detection of the ship target, so before the processing, the template matching method is used to completely separate the sea and land area, and then the ship target is effectively identified by setting the threshold for the pure sea surface and the ship.
Through the process, dozens and hundreds of ship targets in the SAR image can be efficiently detected, and then a user can observe the ship targets in the interested area according to the requirement of the user, so that the working efficiency is greatly improved.
In order to achieve the purpose, the invention is realized by adopting the following technical scheme.
A method of detecting a ship target, the method comprising the steps of:
step 1, obtaining a plurality of historical sea area SAR images and a plurality of historical ship SAR images, wherein each historical sea area SAR image is a sea area amplitude imaging result of a satellite-borne SAR radar on a pure sea area scene, and each historical ship SAR image is a ship amplitude imaging result of the satellite-borne SAR radar on a ship target scene existing on the sea area;
step 2, counting a plurality of sea area amplitude imaging results to obtain the amplitude ranges [ a1, a2] of the sea areas in a plurality of historical sea area SAR images; counting a plurality of ship amplitude imaging results to obtain the amplitude ranges [ b1, b2] of ship targets in a plurality of historical ship SAR images; wherein a2 is less than b 1;
step 3, acquiring a current ship SAR image and an SAR image amplitude matrix corresponding to the current ship SAR image; setting a first detection threshold as c1, and c1 ═ b 1; setting a first marking matrix with the same size as the current ship SAR image and an effectiveness matrix with the same size as the current ship SAR image, and initializing the first marking matrix and the effectiveness matrix to be an all-zero matrix;
step 4, according to the SAR image amplitude matrix and the first detection threshold, strong scattering points are marked in the first marking matrix and the validity matrix;
step 5, setting a second detection threshold as c2, and c2 epsilon [ a2, b 1); setting a second mark matrix with the same size as the current ship SAR image, and initializing the second mark matrix into an all-zero matrix;
step 6, according to the SAR image amplitude matrix and the second detection threshold, strong scattering points and weak scattering points are marked in the second marking matrix;
step 7, rejecting all strong scattering points in the second marker matrix;
step 8, removing isolated strong scattering points in the first marker matrix; screening out strong scattering points belonging to the ship target from the first marker matrix;
and 9, determining the number of the ship targets and the distribution range of each ship target according to the screened strong target points, thereby obtaining the detection result of the ship targets.
The technical scheme of the invention has the characteristics and further improvements that:
(1) in step 2, the amplitude range of the sea area in the plurality of historical sea area SAR images is [0, 50 ]; the amplitude range of pixel points corresponding to the ship targets in the plurality of historical ship SAR images is [150, 1000 ].
(2) The step 4 specifically comprises the following steps:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a first detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the first detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the first marking matrix as 1, and setting the element corresponding to the element position of the SAR image amplitude matrix in the validity matrix as 1, wherein the element marked as 1 corresponds to a strong scattering point.
(3) In step 5, a second detection threshold is set to c2, and c2 ∈ [51, 150).
(4) The step 6 specifically comprises the following steps:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a second detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the second detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the second marking matrix as 1, wherein the element marked as 1 is the sum of the strong scattering point and the weak scattering point.
(5) The step 7 specifically comprises the following steps:
and comparing the elements of the first marker matrix with the elements of the second marker matrix, and setting the value of the element position, corresponding to the element position with 1 in the first marker matrix, in the second marker matrix to be 0, thereby rejecting all strong scattering points in the second marker matrix.
(6) In step 8, removing the isolated strong scattering points in the first marker matrix, specifically including:
setting the resolution of the SAR image as f, setting the side length of a traversal window according to the resolution of the SAR image, wherein the side length of the traversal window is in the range of 20/f to 30/f, traversing all elements in the first marker matrix by using the traversal window, and if the number of the elements with the value of 1 in the traversal window is less than or equal to 2, setting the value of the element of the corresponding first marker matrix in the traversal window as 0, so as to eliminate the isolated strong scattering point in the first marker matrix.
(7) In step 8, screening out strong scattering points belonging to the ship target from the first marker matrix, specifically comprising:
after the isolated strong scattering points in the first marker matrix are eliminated, the remaining strong scattering points in the first marker matrix are marked as actual strong scattering points;
setting the resolution of the SAR image as f, and setting the side length of a first window and the side length of a second window according to the resolution of the SAR image, wherein the side length of the first window is larger than that of the second window; the side length of the first window is in the range of 25 xf to 30 xf, the side length of the second window is in the range of L-20 to L-10, and L represents the side length of the first window;
placing the first window and the second window at a position with a certain actual strong scattering point as a center, and judging whether a weak scattering point exists in an annular area between the first window and the second window;
judging whether a weak scattering point exists in an annular region between the first window and the second window specifically comprises the following steps: if the annular region corresponds to the region in the second marker matrix and has an element of 1, the actual strong scattering point does not belong to a strong scattering point of the ship target (i.e., the actual strong scattering point is a weak scattering point), the element of the first marker matrix corresponding to the actual strong scattering point is set to 0, and the element of the validity matrix corresponding to the actual strong scattering point is set to 0, so that the strong scattering point which is marked as 1 and belongs to the ship target in the first marker matrix is obtained.
(8) The step 9 specifically comprises the following substeps:
(a) setting a ship target number variable, wherein the initial value of the variable is 0;
(b) traversing the first mark matrix, and if the element corresponding to the position in the validity matrix is also 1 for the position with the element of 1 in the first mark matrix, adding 1 to the value of the ship target number variable;
(c) searching other strong scattering points corresponding to the ship target in a preset range by taking the strong scattering point corresponding to the element with the element of 1 in the first marker matrix as a starting point, so as to obtain all the strong scattering points corresponding to the ship target;
(d) setting the element value of the position of each strong scattering point corresponding to the ship target in the validity matrix as 0 for other strong scattering points of the ship target in a preset range;
(e) determining the minimum column number, the maximum column number, the minimum row number and the maximum row number of all strong scattering points of the ship target within a preset range, thereby obtaining the distribution range of the ship target;
(f) and (e) repeatedly executing the substeps (a) to (e) to obtain the total number of the ship targets and the distribution range of each ship target.
Compared with the prior art, the invention has the following advantages: (1) the method has a similar window taking process with the constant false alarm ship detection, and is different in that the constant false alarm detection needs to calculate the mean value and the variance of the taken window in the detection process, the calculated amount is larger, and the calculated amount can be reduced; (2) when the method is used for distinguishing land areas and sea areas, whether the annular area between the two windows is 1 or not can effectively distinguish strong scattering points. (3) The method can effectively separate the ocean area from the wharf area without the help of the prior information of the port, reduces the dependence on the prior data, and avoids the error caused by the low precision of the prior information.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a method for detecting a ship target according to an embodiment of the present invention;
fig. 2 is a schematic process diagram for determining the number of ship targets and the distribution range of each ship target according to the embodiment of the present invention;
fig. 3 is a schematic diagram of a high resolution third SAR image of a Ningbo bay port with a resolution of 5m according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a binary image of a ningbo bay port SAR image larger than a first threshold according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a detection result of a ship target on a small part of the sea area in the SAR image according to the embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides a detection method of a ship target, and with reference to fig. 1, the method comprises the following steps:
the method comprises the following steps of 1, obtaining a plurality of historical sea area SAR images and a plurality of historical ship SAR images, wherein each historical sea area SAR image is a sea area amplitude imaging result of a satellite-borne SAR radar on a pure sea area scene, and each historical ship SAR image is a ship amplitude imaging result of the satellite-borne SAR radar on a ship target scene existing on the sea area.
Step 2, counting a plurality of sea area amplitude imaging results to obtain the amplitude ranges [ a1, a2] of the sea areas in a plurality of historical sea area SAR images; counting a plurality of ship amplitude imaging results to obtain the amplitude ranges [ b1, b2] of the corresponding pixels of the ships in a plurality of historical ship SAR images; wherein a2 < b 1.
In step 2, the amplitude range of the pixels corresponding to the sea area parts in the plurality of historical sea area SAR images is [0, 50 ]; the amplitude range of the ship-corresponding pixels in the plurality of historical ship SAR images is [150, 1000 ].
Specifically, the backscattering of electromagnetic waves caused by the specular reflection of the sea plane is very weak, the echo intensity of the sea radar is very weak due to the low backscattering coefficient, the sea plane is in a continuous motion state, the sea plane is seriously decohered due to the fluctuation of sea waves, and the gain of a sea area target is also reduced. Therefore, according to the characteristic of the sea, firstly, the amplitude values of the sea areas in tens of sea area SAR images are observed and counted, the amplitude values of the sea areas in the SAR images are concentrated in the interval of [0, 50], the values have no dimension, the value represents the strength information of the images, the value mainly depends on the backscattering coefficient of the ground objects, the larger the backscattering coefficient is, the brighter the backscattering coefficient is in the SAR images, the smaller the backscattering coefficient is, the darker the backscattering coefficient is in the SAR images, and the sea surface belongs to the area with the very small backscattering coefficient.
Similarly, observation and statistics are carried out on the ship amplitude in the ten sea area SAR images, and the ship amplitude values in the SAR images are found to be concentrated in the interval of [150, 1000], a ship target belongs to a target with a large backscattering coefficient, the backscattering coefficient is related to the material of the target, and the backscattering coefficients of metal targets are large. Therefore, for the detection of the ship target on the sea surface, a 150 magnitude threshold can be set according to the empirical value, and is recorded as a first detection threshold.
Step 3, acquiring a current ship SAR image and an SAR image amplitude matrix corresponding to the current ship SAR image; setting a first detection threshold as c1, and c1 ═ b 1; setting a first marking matrix with the same size as the current ship SAR image and a validity matrix with the same size as the current ship SAR image, and initializing the first marking matrix, wherein the validity matrix is an all-zero matrix.
And 4, marking strong scattering points in the first marking matrix and the validity matrix according to the SAR image amplitude matrix and the first detection threshold.
The step 4 specifically comprises the following steps:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a first detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the first detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the first marking matrix as 1, and setting the element corresponding to the element position of the SAR image amplitude matrix in the validity matrix as 1, wherein the element marked as 1 corresponds to a strong scattering point.
Step 5, setting a second detection threshold as c2, and c2 epsilon [ a2, b 1); and setting a second marking matrix with the same size as each historical sea area SAR image, and initializing the second marking matrix into an all-zero matrix.
In step 5, a second detection threshold is set to c2, and c2 ∈ [51, 150).
And 6, marking strong scattering points and weak scattering points in the second marking matrix according to the SAR image amplitude matrix and the second detection threshold.
The step 6 specifically comprises the following steps:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a second detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the second detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the second marking matrix as 1, wherein the element marked as 1 is the sum of the strong scattering point and the weak scattering point.
And 7, rejecting all strong scattering points in the second marker matrix.
The step 7 specifically comprises the following steps:
and comparing the elements of the first marker matrix with the elements of the second marker matrix, and setting the value of the element position, corresponding to the element position with 1 in the first marker matrix, in the second marker matrix to be 0, thereby rejecting all strong scattering points in the second marker matrix.
The steps 3 to 7 are specifically as follows: two matrixes with the same size as the SAR image, namely a marking matrix 1(F1ag1 matrix) and a validity matrix (vaild matrix), are created and initialized to be an all-zero matrix, the marking matrix 1 is denoted as F1, and the validity matrix is denoted as V. And traversing the SAR image amplitude matrix, recording the coordinates of pixels with amplitude values larger than a first detection threshold in the SAR image, and setting the values of the matrix F1 and the corresponding positions in the matrix V as 1.
A new flag matrix 2(flag2 matrix) is created, initialized to an all-zero matrix, and the flag matrix 2 is used to mark weak objects, which will be denoted as F2 hereinafter. In addition, a new detection threshold is set, and this threshold should be set to [51, 150 ] according to experience, and is not set to 75, and is recorded as the second detection threshold. And traversing the SAR image amplitude matrix, recording the coordinates of pixels with amplitude values larger than a second detection threshold in the SAR image, and setting the value of the corresponding position in the matrix F2 as 1. Now, the matrix F2 includes a strong target point and a weak target point, the coordinates of the pixels of which the values in the tag matrix 1 and the tag matrix 2 are both 1 are counted, and the value of the corresponding position in the tag matrix 2 is set to 0, that is, the strong target point is rejected.
Considering that the length of ships is typically greater than 50 meters, even in a 5 meter resolution SAR image, a ship has a length of more than 10 pixels, so isolated strong scattering points in the labeling matrix F1 do not belong to ship targets. The number of strong scattering points in a window of 5 × 5 (pixels) is counted by traversing the label matrix F1, and if the number of pixels with a value of 1 is less than or equal to 2, all values in the window are set to 0, i.e., the isolated strong scattering points are rejected. The selection of the window size is related to the resolution of the SAR image, and for the SAR image with the resolution of x meters, the value range of the window size can be selected from 20/x to 30/x.
Step 8, removing isolated strong scattering points in the first marker matrix; and screening out strong scattering points belonging to the ship target from the first marker matrix.
In step 8, removing the isolated strong scattering points in the first marker matrix, specifically including:
setting the resolution of the SAR image as f, setting the side length of a traversal window according to the resolution of the SAR image, wherein the side length of the traversal window is in the range of 20/f to 30/f, traversing all elements in the first marker matrix by using the traversal window, and if the number of the elements with the value of 1 in the traversal window is less than or equal to 2, setting the value of the element of the corresponding first marker matrix in the traversal window as 0, so as to eliminate the isolated strong scattering point in the first marker matrix.
In step 8, screening out strong scattering points belonging to the ship target from the first marker matrix, specifically comprising:
after the isolated strong scattering points in the first marker matrix are eliminated, the remaining strong scattering points in the first marker matrix are marked as actual strong scattering points;
setting the resolution of the SAR image as f, and setting the side length of a first window and the side length of a second window according to the resolution of the SAR image, wherein the side length of the first window is larger than that of the second window; the side length of the first window is in the range of 25 xf to 30 xf, the side length of the second window is in the range of L-20 to L-10, and L represents the side length of the first window;
placing the first window and the second window at a position with a certain actual strong scattering point as a center, and judging whether a weak scattering point exists in an annular area between the first window and the second window;
if the annular area corresponds to the area in the second marker matrix, the element is 1, the actual strong scattering point does not belong to the strong scattering point of the ship target, the element of the first marker matrix corresponding to the actual strong scattering point is set to 0, and the element of the validity matrix corresponding to the actual strong scattering point is set to 0, so that the strong scattering point which is marked as 1 and belongs to the ship target in the first marker matrix is obtained.
As already mentioned in the foregoing analysis, since the electromagnetic wave irradiated on the sea surface mainly undergoes specular reflection, the echo signal is weak; electromagnetic waves irradiate the ship target or land areas and mainly diffuse reflection occurs, so the ship target has the following characteristics in the SAR image: the amplitude of the ship target is very strong, the signal amplitude in a certain range around the ship target is very weak, and the strong target in the land area does not have the characteristic. It is based on this feature to propose a detection model taking a large window, e.g. 150 x 150 pixels, centered on the strong scattering point (this parameter applies to SAR images with a resolution of 5 meters). And taking a small window of 130 x 130 pixels by taking the strong scattering point as the center, and then counting whether a weak target signal exists in an annular area between the two windows. The selection of the window size is related to the resolution of the SAR image, and for the SAR image with the resolution of x meters, the value range of the side length L of the large window is as follows: 25 x-30 x, and the side length S of the small window ranges from L-20 to L-10. If a weak target signal is found in the annular area corresponding to the marker matrix F2, the strong scattering point does not belong to the sea surface ship target, and the strong scattering point is removed from the marker matrix F1 and the validity matrix V, namely the value of the scattering point is changed from 1 to 0.
And 9, determining the number of the ship targets and the distribution range of each ship target according to the screened strong target points, thereby obtaining the detection result of the ship targets.
The step 9 specifically comprises the following substeps:
(a) setting a ship target number variable, wherein the initial value of the variable is 0;
(b) traversing the first mark matrix, and if the element corresponding to the position in the validity matrix is also 1 for the position with the element of 1 in the first mark matrix, adding 1 to the value of the ship target number variable;
(c) searching other strong scattering points corresponding to the ship target in a preset range by taking the strong scattering point corresponding to the element with the element of 1 in the first marker matrix as a starting point, so as to obtain all the strong scattering points corresponding to the ship target;
(d) setting the element value of the position of each strong scattering point corresponding to the ship target in the validity matrix as 0 for other strong scattering points of the ship target in a preset range;
(e) determining the minimum column number, the maximum column number, the minimum row number and the maximum row number of all strong scattering points of the ship target within a preset range, thereby obtaining the distribution range of the ship target;
(f) and (e) repeatedly executing the substeps (a) to (e) to obtain the total number of the ship targets and the distribution range of each ship target.
Through the screening processing in step 9, only the positions of the strong scattering points on the sea surface ship targets are obtained preliminarily, and how many sea surface ship targets exist in the SAR image and the distribution range of each ship target need to be further determined.
Specifically, the label matrix F1 is traversed, and after a strong scattering point is encountered, it is determined whether the pixel in the validity matrix V is 1. If the value of the scattering point in the validity matrix is 1, indicating that the scattering point has not been processed and belongs to a new ship target, then adding 1 to the number of ship targets, and then searching other scattering points of the ship target in three directions of downward, leftward and rightward by using the scattering point as a starting point (no upward search is needed because the traversal order is from bottom to top, and the upper strong scattering points have been processed), as shown in fig. 2, a schematic process diagram for determining the number of ship targets and the distribution range of each ship target provided by the embodiment of the present invention is shown:
and (3) searching 150 rows downwards from the starting point, searching 75 columns leftwards and 75 columns rightwards in each row of data, and if a strong scattering point is encountered, namely the value of the pixel point in the marker matrix is 1, setting the value of the corresponding pixel in the validity matrix to be 0 because the corresponding pixel and the starting point belong to the same ship target instead of a new ship target. Counting the minimum column number and the maximum column number of the strong scattering point on the ship target so as to determine the left and right boundaries of the ship target, wherein the upper boundary can be determined according to the row number of the starting point, and the determination method of the lower boundary comprises the following steps: if no strong scattering point is detected in the continuous 5 rows of data, the lower boundary of the ship target is considered to be searched, and the 150 rows do not need to be searched continuously, so that two adjacent ship targets are prevented from being mistakenly judged as one detection target. Thus, the distribution range of the ship target is determined. After traversing the mark matrix once, the number of the sea surface ships and the distribution range of each ship target can be obtained.
Simulation experiment:
fig. 3 (high-resolution three-size SAR image at 5m resolution at gulf harbor) illustrates how ship identification is performed using two windows of different sizes. The rectangular frame part is used to show how the two marker matrices achieve the purpose of detection.
Fig. 4 (binary image of the nivewan port SAR image larger than the first threshold) illustrates that all points in the SAR image whose amplitude is larger than the second detection threshold are set to 1 larger than the threshold and set to 0 smaller than the threshold, so that the label matrix F2 is obtained. Wherein, the obvious difference between the strong scattering points on the land and the sea surface ships can be obviously seen, namely, the annular area between the first window (150 x 150) and the second window (130 x 130) has no point with 1 for the sea surface, and more or less points with 1 for the land, so that the strong scattering points on the sea surface and the land can be effectively distinguished.
Through the screening processing, the position of the strong scattering point on the sea surface ship target is obtained preliminarily. After the ship target is identified, the area where the ship is located is marked out by using a circular frame in order to facilitate observation of the position of the target, so that the ship target is identified conveniently.
Fig. 5 shows a small part of the sea area in the SAR image, and shows the effect of ship detection. The ship target detected in the graph is marked by a circular frame.
Through calculation, the detection accuracy is over 90%. And the whole algorithm program runs a satellite-borne SAR image of 20000 multiplied by 16000, the time consumption is only 3 minutes, and the efficiency is quite high.
Compared with the prior art, the invention has the following advantages: (1) the method has a similar window-taking process with the constant false alarm ship detection, and is different in that the constant false alarm detection needs to calculate the mean value and the variance of the window-taking in the detection process, the calculated amount is large, and the calculated amount can be reduced. (2) When the method is used for distinguishing land areas and sea areas, whether the annular area between the two windows is 1 or not can effectively distinguish strong scattering points.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (9)
1. A method of detecting a ship target, the method comprising the steps of:
step 1, obtaining a plurality of historical sea area SAR images and a plurality of historical ship SAR images, wherein each historical sea area SAR image is a sea area amplitude imaging result of a satellite-borne SAR radar on a pure sea area scene, and each historical ship SAR image is a ship amplitude imaging result of the satellite-borne SAR radar on a ship target scene existing on the sea area;
step 2, counting a plurality of sea area amplitude imaging results to obtain the amplitude ranges [ a1, a2] of the sea areas in a plurality of historical sea area SAR images; counting a plurality of ship amplitude imaging results to obtain the amplitude ranges [ b1, b2] of ship targets in a plurality of historical ship SAR images; wherein a2 is less than b 1;
step 3, acquiring a current ship SAR image and an SAR image amplitude matrix corresponding to the current ship SAR image; setting a first detection threshold as c1, and c1 ═ b 1; setting a first marking matrix with the same size as the current ship SAR image and an effectiveness matrix with the same size as the current ship SAR image, and initializing the first marking matrix and the effectiveness matrix to be an all-zero matrix;
step 4, according to the SAR image amplitude matrix and the first detection threshold, strong scattering points are marked in the first marking matrix and the validity matrix;
step 5, setting a second detection threshold as c2, and c2 epsilon [ a2, b 1); setting a second mark matrix with the same size as the current ship SAR image, and initializing the second mark matrix into an all-zero matrix;
step 6, according to the SAR image amplitude matrix and the second detection threshold, strong scattering points and weak scattering points are marked in the second marking matrix;
step 7, rejecting all strong scattering points in the second marker matrix;
step 8, removing isolated strong scattering points in the first marker matrix; screening out strong scattering points belonging to the ship target from the first marker matrix;
and 9, determining the number of the ship targets and the distribution range of each ship target according to the screened strong scattering points belonging to the ship targets, thereby obtaining the detection result of the ship targets.
2. The method for detecting the ship target according to claim 1, wherein in step 2, the range of the sea area in the plurality of historical sea area SAR images is [0, 50 ]; the amplitude range of the ship target in the plurality of historical ship SAR images is [150, 1000 ].
3. The method for detecting the ship target according to claim 1, wherein the step 4 specifically comprises:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a first detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the first detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the first marking matrix as 1, and setting the element corresponding to the element position of the SAR image amplitude matrix in the validity matrix as 1, wherein the element marked as 1 corresponds to a strong scattering point.
4. The method for detecting the ship target according to claim 1, wherein in step 5, the second detection threshold is set to be c2, and c2 e [51, 150 ].
5. The method for detecting the ship target according to claim 1, wherein the step 6 specifically comprises:
traversing the SAR image amplitude matrix, and judging the size of each element in the SAR image amplitude matrix and a second detection threshold;
if the value of a certain element in the SAR image amplitude matrix is larger than the second detection threshold, setting the element corresponding to the element position of the SAR image amplitude matrix in the second marking matrix as 1, wherein the element marked as 1 is the sum of the strong scattering point and the weak scattering point.
6. The method for detecting the ship target according to claim 1, wherein the step 7 specifically comprises:
and comparing the elements of the first marker matrix with the elements of the second marker matrix, and setting the value of the element position, corresponding to the element position with 1 in the first marker matrix, in the second marker matrix to be 0, thereby rejecting all strong scattering points in the second marker matrix.
7. The method for detecting ship targets according to claim 1, wherein the step 8 of rejecting isolated strong scattering points in the first marker matrix specifically comprises:
setting the resolution of the SAR image as f, setting the side length of a traversal window according to the resolution of the SAR image, wherein the side length of the traversal window is in the range of 20/f to 30/f, traversing all elements in the first marker matrix by using the traversal window, and if the number of the elements with the value of 1 in the traversal window is less than or equal to 2, setting the value of the element of the corresponding first marker matrix in the traversal window as 0, so as to eliminate the isolated strong scattering point in the first marker matrix.
8. The method for detecting ship targets according to claim 1, wherein in step 8, the step of screening out strong scattering points belonging to the ship targets from the first marker matrix specifically comprises:
after the isolated strong scattering points in the first marker matrix are eliminated, the remaining strong scattering points in the first marker matrix are marked as actual strong scattering points;
setting the resolution of the SAR image as f, and setting the side length of a first window and the side length of a second window according to the resolution of the SAR image, wherein the side length of the first window is larger than that of the second window; the side length of the first window is in the range of 25 xf to 30 xf, the side length of the second window is in the range of L-20 to L-10, and L represents the side length of the first window;
placing the first window and the second window at a position with a certain actual strong scattering point as a center, and judging whether a weak scattering point exists in an annular area between the first window and the second window;
judging whether a weak scattering point exists in an annular region between the first window and the second window specifically comprises the following steps: if the annular area corresponds to the area in the second marker matrix, the element is 1, the actual strong scattering point does not belong to the strong scattering point of the ship target, the element of the first marker matrix corresponding to the actual strong scattering point is set to 0, and the element of the validity matrix corresponding to the actual strong scattering point is set to 0, so that the strong scattering point which is marked as 1 and belongs to the ship target in the first marker matrix is obtained.
9. The method for detecting the ship target according to claim 1, wherein the step 9 specifically comprises the following substeps:
(a) setting a ship target number variable, wherein the initial value of the variable is 0;
(b) traversing the first mark matrix, and if the element corresponding to the position in the validity matrix is also 1 for the position with the element of 1 in the first mark matrix, adding 1 to the value of the ship target number variable;
(c) searching other strong scattering points corresponding to the ship target in a preset range by taking the strong scattering point corresponding to the element with the element of 1 in the first marker matrix as a starting point, so as to obtain all the strong scattering points corresponding to the ship target;
(d) setting the element value of the position of each strong scattering point corresponding to the ship target in the validity matrix as 0 for other strong scattering points of the ship target in a preset range;
(e) determining the minimum column number, the maximum column number, the minimum row number and the maximum row number of all strong scattering points of the ship target within a preset range, thereby obtaining the distribution range of the ship target;
(f) and (e) repeatedly executing the substeps (a) to (e) to obtain the total number of the ship targets and the distribution range of each ship target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711049084.8A CN107862271B (en) | 2017-10-31 | 2017-10-31 | Detection method of ship target |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711049084.8A CN107862271B (en) | 2017-10-31 | 2017-10-31 | Detection method of ship target |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107862271A CN107862271A (en) | 2018-03-30 |
| CN107862271B true CN107862271B (en) | 2021-05-25 |
Family
ID=61697966
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711049084.8A Active CN107862271B (en) | 2017-10-31 | 2017-10-31 | Detection method of ship target |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107862271B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108694714A (en) * | 2018-05-14 | 2018-10-23 | 浙江大学 | Ship seakeeping system in a kind of adaptive colony intelligence optimization SAR Radar Seas |
| CN108629382A (en) * | 2018-05-14 | 2018-10-09 | 浙江大学 | Ship seakeeping system in a kind of gunz optimization SAR Radar Seas |
| CN109977892B (en) * | 2019-03-31 | 2020-11-10 | 西安电子科技大学 | Ship detection method based on local saliency features and CNN-SVM |
| CN111680565B (en) * | 2020-05-08 | 2022-06-07 | 湖北航天技术研究院总体设计所 | Port area ship target detection method based on SAR image |
| CN113820712B (en) * | 2021-09-07 | 2023-07-28 | 中山大学 | Ship target positioning method and system based on strong scattering points |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102144173A (en) * | 2008-05-29 | 2011-08-03 | 电视广播有限公司 | Target detection in a sar-imaged sea area |
| CN103164686A (en) * | 2011-12-12 | 2013-06-19 | 中国科学院对地观测与数字地球科学中心 | Method for identifying ship type according to high resolution synthetic aperture radar (SAR) image of ship |
| CN103400156A (en) * | 2013-07-04 | 2013-11-20 | 西安电子科技大学 | CFAR (Constant False Alarm Rate) and sparse representation-based high-resolution SAR (Synthetic Aperture Radar) image ship detection method |
| CN103558593A (en) * | 2013-10-16 | 2014-02-05 | 北京航空航天大学 | Method for obtaining position of vessel particular point based on strong scattering points |
| CN104077777A (en) * | 2014-07-04 | 2014-10-01 | 中国科学院大学 | Sea surface vessel target detection method |
| CN104091335A (en) * | 2014-07-04 | 2014-10-08 | 西安电子科技大学 | Polarization SAR image ship target detection method |
| CN104166838A (en) * | 2014-07-16 | 2014-11-26 | 西安电子科技大学 | Multi-layer CFAR-based layered high-resolution SAR image ship detection method |
| CN104376330A (en) * | 2014-11-19 | 2015-02-25 | 西安电子科技大学 | Polarization SAR image ship target detection method based on superpixel scattering mechanism |
-
2017
- 2017-10-31 CN CN201711049084.8A patent/CN107862271B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102144173A (en) * | 2008-05-29 | 2011-08-03 | 电视广播有限公司 | Target detection in a sar-imaged sea area |
| CN103164686A (en) * | 2011-12-12 | 2013-06-19 | 中国科学院对地观测与数字地球科学中心 | Method for identifying ship type according to high resolution synthetic aperture radar (SAR) image of ship |
| CN103400156A (en) * | 2013-07-04 | 2013-11-20 | 西安电子科技大学 | CFAR (Constant False Alarm Rate) and sparse representation-based high-resolution SAR (Synthetic Aperture Radar) image ship detection method |
| CN103558593A (en) * | 2013-10-16 | 2014-02-05 | 北京航空航天大学 | Method for obtaining position of vessel particular point based on strong scattering points |
| CN104077777A (en) * | 2014-07-04 | 2014-10-01 | 中国科学院大学 | Sea surface vessel target detection method |
| CN104091335A (en) * | 2014-07-04 | 2014-10-08 | 西安电子科技大学 | Polarization SAR image ship target detection method |
| CN104166838A (en) * | 2014-07-16 | 2014-11-26 | 西安电子科技大学 | Multi-layer CFAR-based layered high-resolution SAR image ship detection method |
| CN104376330A (en) * | 2014-11-19 | 2015-02-25 | 西安电子科技大学 | Polarization SAR image ship target detection method based on superpixel scattering mechanism |
Non-Patent Citations (3)
| Title |
|---|
| A ship detector using invariant scattering feature for polarimetric SAR images;Yuan SUN et al.;《IGARSS 2012》;20121231;第6817-6820页 * |
| Study on One-dimensional Range Profile Recognition Based on Scattering Features;Meng-da CUI et al.;《2014 International Conference on Electrical and Electronic Engineering (EEE 2014)》;20141231;第304-309页 * |
| 大场景SAR图像舰船目标快速检测;甄勇 等;《信息工程大学学报》;20161031;第17卷(第5期);第535-540页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107862271A (en) | 2018-03-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107862271B (en) | Detection method of ship target | |
| CN109583293B (en) | Aircraft target detection and identification method in satellite-borne SAR image | |
| Khosravipour et al. | Generating pit-free canopy height models from airborne lidar | |
| CN110031843B (en) | ROI (region of interest) -based SAR (synthetic Aperture Radar) image target positioning method, system and device | |
| Pitkänen et al. | Adaptive methods for individual tree detection on airborne laser based canopy height model | |
| US8154952B1 (en) | Method and system for real-time automated change detection and classification for images | |
| CN109471098B (en) | Airport runway foreign matter detection method utilizing FOD radar phase coherence information | |
| CN110443201B (en) | Target identification method based on multi-source image joint shape analysis and multi-attribute fusion | |
| CN109829423A (en) | A kind of icing lake infrared imaging detection method | |
| Song et al. | Efficient water area classification using Radarsat-1 SAR imagery in a high relief mountainous environment | |
| Isiacik Colak et al. | Coastline zone extraction using Landsat-8 OLI imagery, case study: Bodrum Peninsula, Turkey | |
| CN107835998A (en) | For identifying the layering Tiling methods of the surface type in digital picture | |
| CN111105419B (en) | Vehicle and ship detection method and device based on polarized SAR image | |
| CN114545410B (en) | Crop lodging monitoring method based on synthetic aperture radar dual-polarized data coherence | |
| Angelini et al. | A review and test of shoreline extraction techniques | |
| Zhao et al. | Automatic extraction of floating ice at Antarctic continental margin from remotely sensed imagery using object-based segmentation | |
| CN113343819A (en) | Efficient unmanned aerial vehicle-mounted SAR image target segmentation method | |
| Febriawan et al. | Detection and characterization of an archaeological wreck site in Sunda Strait, Indonesia | |
| Dang et al. | Discussion on sea ice segmentation of high resolution radar data | |
| Wang et al. | Real-time extraction of water surface boundary using shipborne radar | |
| Bae et al. | Land Masking Methods of Sentinel-1 SAR Imagery for Ship Detection Considering Coastline Changes and Noise | |
| Tian | Automatic target detection and analyses in side-scan sonar imagery | |
| Busko et al. | Preprocessing of SAR Image for Building Detection | |
| Xu et al. | On-borne automated detection of ocean internal waves in SAR images based on spatial features | |
| Komura et al. | Analysis of size of tree crown on aerial image using circle unification |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Li Zhenfang Inventor after: Xing Chao Inventor after: Tian Feng Inventor after: Mao Qin Inventor after: Wang Zhibin Inventor before: Li Zhenfang Inventor before: Xing Chao Inventor before: Tian Feng Inventor before: Mao Qin |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |