WO2021051848A1

WO2021051848A1 - Radar effective detection area extraction method based on remote sensing image

Info

Publication number: WO2021051848A1
Application number: PCT/CN2020/091694
Authority: WO
Inventors: 徐凯; 满青珊; 王君; 隋远; 郑浩; 周士胜
Original assignee: 南京莱斯网信技术研究院有限公司
Priority date: 2019-09-17
Filing date: 2020-05-22
Publication date: 2021-03-25
Also published as: CN110568416B; CN110568416A

Abstract

The present invention provides a radar effective detection area extraction method based on a remote sensing image. A radar interference area and an effective detection area are extracted, a false object caused by ground clutter is eliminated, offshore location distribution information of a radar target is visually and truly reflected, and a real basis is provided for traffic situation evaluation. The method is different from other common interference area extraction methods or effective detection area extraction methods, a remote sensing image is adopted to assist in distinguishing the type of an area of a radar detection range, an interference area (an area generated by clutters such as land and islands) and an effective detection area are accurately extracted, the accuracy degrees of detection and entry are improved, and a reliable basis is provided for radar target display of a VTS system.

Description

An Effective Radar Detection Area Extraction Method Based on Remote Sensing Image

Technical field

The invention relates to a method for extracting a radar effective detection area based on remote sensing images.

Background technique

Radar target detection technology is an important technology used in navigation and water traffic supervision. It is important for improving the VTS (VTS, vessel traffic service, meaning ship traffic service) system’s ability to collect dynamic information on ships, and to ensure the safety of marine navigation and water transportation. Traffic order plays an important role. With the increase of human commercial production activities, the number of ships navigating in major ports and waterways is increasing. Radar target detection provides important reference value for water traffic supervision, but there are often ground clutters during radar detection. It interferes with the detection and admission of targets, so the radar area detection map is made in advance, the water surface area is designated as the effective radar detection area, the land area is removed, the false target is reduced, and the error similar to the ship traveling to the land is eliminated.

Currently, the two most common effective area extraction methods are the nautical chart detection area extraction method and the manual detection area extraction. The method of extracting the detection area of the nautical chart is to use the water depth information of the nautical chart, traverse the scanning area of the radar point by point, extract the water depth information of the corresponding position, and filter the detection area of the radar; manual extraction is to manually draw the land area in conjunction with the nautical chart. Make the land area of the radar, and then traverse the radar scanning area point by point to judge the location point information, and the location that is not in the drawn land area is the final detection area. The above two methods have corresponding shortcomings respectively. The method of extracting the detection area of the sea chart is based on the electronic chart data. The electronic chart has the disadvantages of slow updating, some area data does not conform to the actual situation, and the high price of updating the chart; manual drawing The detection area method requires manual drawing of the land area. Manual drawing has errors, and the map details are difficult to control. When the detection area is too large, there will be a problem of low extraction efficiency.

Summary of the invention

In view of the shortcomings of the prior art, the present invention provides a method for extracting effective radar detection areas based on remote sensing images, which includes the following steps:

Step 1. Set the parameters, use the electronic chart to superimpose the remote sensing image data, and use the water depth information of the electronic chart combined with the edge detection method to generate the detection information index map;

Step 2: Traverse all the positions (polar coordinates) of the radar scanning area, and perform polar coordinate conversion on each position point to obtain the detection information of the position point. The detection information is the detection information index map generated in step 1:

Step 3: Judge the detection information, determine whether the location point is a land area, and record the area category information of the location point;

Step 4. Perform the output file operation to complete the extraction of the effective detection area.

Step 1 includes the following steps:

Step 1-1, set the parameters: set the radar center, scan radius, azimuth value and sampling accuracy, based on the water depth information provided by the electronic chart, extract the water depth information in a certain range centered on the radar, and the water depth is greater than 0 If it is a marine area, otherwise it is a land area. A map layer containing sea and land boundary lines is obtained to generate a land and sea map. All the map sheets containing sea and land boundary lines in the electronic chart layer can be extracted and saved for subsequent use;

Step 1-2: Obtain remote sensing images of the same or similar scale for the map layer containing the sea and land boundary lines, perform gray-scale processing on them to obtain gray-scale images, and then perform edge extraction on the gray-scale images. The gradient operator is:

In the gradient operator, F(i,j) represents the gray value of the image, i represents the abscissa of the image, and j represents the ordinate of the image. This operator can exactly reflect the grayscale changes on the edge of the image. After the edge detection is completed The sea and land boundaries can be displayed more clearly. This step corrects the coastline error caused by the electronic chart may not be updated for a long time, and then normalizes the size of the image containing the accurate sea and land boundary information and updates it back to the corresponding position.

Steps 1-3, after updating the sea and land boundary line as described in the above step 1-2, all detection information is generated, and the closed area block is divided according to the sea and land boundary line and the circular area line scanned by the radar, and whether the water depth is greater than The proportion in the zero statistical area block. If the proportion of the water depth is greater than zero is greater than 50%, it is determined to be a land area and the detection information is 1, otherwise it is determined to be an ocean area and the detection information is 0, thereby generating a complete detection information index map.

Step 2 includes the following steps:

Step 2-1, calculate the average distance represented by the unit longitude and unit latitude;

Step 2-2, convert the polar coordinates of each location point into latitude and longitude coordinates;

In step 2-3, the detection information index map generated in step 1 is retrieved according to the latitude and longitude coordinates to obtain the detection information detection data.

Step 2-1 includes:

In the current display range of the electronic chart, take the left middle point and the right middle point in the horizontal direction, record the longitude difference between the two points, and then call the longitude and latitude-distance calculation method of the electronic chart to calculate the distance between the two points Then divide the distance difference by the longitude difference to get the average distance represented by the unit longitude; in the current display range of the electronic chart, take the top middle point and the bottom middle point in the vertical direction as represented by the calculation unit longitude The average distance is calculated in the same way as the average distance represented by the unit latitude. .

Step 2-2 includes: approximate the earth as a sphere, the position point is centered on the radar, and the scanning radius is taken as the radius of the circular area. The polar coordinate information of the position point includes the azimuth value offset and the position point serial number, which is set in step 1. After calculating the center latitude and longitude, scanning radius, sampling accuracy, and azimuth value of the radar, first calculate the upper limit d of the position number sequence number:

Then calculate the spherical distance dis between the corresponding position point and the radar center, and the calculation formula is:

dis=c*pno,

pno∈[0,d),

Then obtain the angle information ang from the azimuth offset of the position point:

offset∈[0,P),

According to the trigonometric function conversion, calculate the distance dis_x from the radar center point in the longitude direction that is the horizontal direction and the distance dis_y from the radar center point in the latitude direction that is the vertical direction:

dis_x=dis*sinang,

dis_y=dis*cosang,

According to the average distance represented by the unit longitude and latitude calculated in step 2-1, the longitude value lon and latitude value lat of the corresponding position are calculated, and the formula is as follows:

lon=dlon+dis_x/dx,

lat=dlat+dis_y/dy,

Among them, dlon, dlat respectively represent the longitude and latitude of the radar center, dx, dy respectively represent the average distance per unit longitude and the average distance per unit latitude solved in step 2-1, P represents the azimuth value (constant), and l represents the scanning radius. c represents the sampling accuracy (unit: meter), the parameters P, l, c, dlon, and dlat in step 1 are determined by setting, offset represents the offset of the azimuth value, and pno represents the sequence number of the position point, which belongs to polar coordinate information.

Step 3: According to the principle of generating the detection information index map in step 1, it is judged whether the location point is a water surface area, if it is, the area category result corresponding to the location point is 0, otherwise it is 1.

Step 4 includes: compose the location points whose area category result is 1 into the land area of the detection map, and the location points whose result is 0 form the water area of the detection map. The water area of the detection map is the effective detection area and displayed in the VTS system. Inspection chart.

In the circular radar scanning area, the method of the present invention determines the number of polar coordinates according to the unit of offsetting the true north direction of the radar center point and the distance from the radar center point, and then converts the polar coordinates into longitude and latitude point information point by point to convert the position The information is passed in as a parameter, and the detection information generated based on remote sensing images, sea chart water depth data and edge detection technology is used to determine the area type of the point (whether it belongs to land), and finally output to the detection map file according to the discrimination result to complete the effective detection area Extraction.

Beneficial effects: The present invention proposes an automatic radar effective area extraction method based on remote sensing images. In the scanning area of the radar, GIS software is used to superimpose the sea chart and remote sensing image data, and the water depth information is used to preliminarily determine the boundary line of the sea and land, and then According to the location of the sea and land boundary, extract small-scale remote sensing images at the same or similar scale, use edge detection technology to update the location of the sea and land boundary, generate a detection information index map, and convert the radar's polar coordinates into latitude and longitude coordinates to find the corresponding location Point, extract the corresponding detection information, determine whether the point is a land area, and then traverse point by point, generate a detection map of the entire scanning area, and extract the effective radar detection area. Using this method, there is no need to manually draw the land area and avoid errors caused by manual drawing. At the same time, remote sensing images and edge detection technology are used to improve the accuracy of water depth information to determine the sea and land boundary, and the effective detection area is automatically extracted, which is more efficient than manual methods. High, the update of remote sensing images is faster than the update of electronic nautical charts, avoiding the disadvantages of slower and expensive update of nautical charts, and is more realistic and environmental changes, and the accuracy and accuracy are guaranteed.

The present invention aims at the advantages and disadvantages of the current two kinds of radar effective area extraction, comprehensive remote sensing image, edge detection technology and their characteristics, and proposes the present invention, which takes into account the advanced nature of remote sensing image and the advantages of automatically extracting the detection area, and avoids manual drawing of land. Accurately and quickly extracting the effective detection area of the radar, etc., the detail error caused by the area.

Description of the drawings

In the following, the present invention will be further described in detail with reference to the accompanying drawings and specific embodiments, and the above-mentioned and/or other advantages of the present invention will become clearer.

Figure 1 is a flowchart of a method for extracting effective radar detection areas based on remote sensing images.

Figure 2 is an area detection map based on remote sensing images.

Figure 3 is a radar detection effect diagram of unused area detection technology.

Figure 4 is a radar detection effect diagram using area detection technology.

detailed description

As shown in Fig. 1, the present invention provides a method for extracting a radar effective detection area based on remote sensing images, which includes the following steps:

Step 1-1, set the radar's position information (dlon, dlat), determine its scanning radius l, sampling accuracy c, and azimuth value P. The mechanical radar commonly used in VTS usually sets the scanning radius to 37.5km, the sampling accuracy to 15m, and the azimuth value. Take 8192, so 8192*(37.5*1000/15)=20480000 positions will be identified. At the same time, GIS software is used to superimpose electronic chart and remote sensing image data. Based on the water depth information provided by the electronic chart, the water depth information is extracted from a rectangular area with the radar as the center and the side length as the scanning diameter. The water depth is greater than 0 as the ocean area. , Otherwise it is a land area, generate a land and sea map, and all the map sheets containing the sea and land boundary lines in the electronic chart layer can be extracted and saved for subsequent use.

Step 1-2, the previous step generates the sea-land map containing the sea-land boundary line. For the map layer containing the sea-land boundary line, obtain the remote sensing image of the same or similar scale, and perform the gray-scale processing on it to obtain the gray-scale image. , And then perform edge extraction on the gray image, the gradient operator for edge extraction is:

In the gradient operator, F(i,j) represents the gray value of the image, i represents the abscissa of the image, and j represents the ordinate of the image. This operator can exactly reflect the grayscale changes on the edge of the image. After the edge detection is completed It can display the sea and land boundary more clearly, normalize the size of the image containing accurate boundary information according to the scale, latitude and longitude information, and superimpose and update the original position.

Steps 1-3, after updating the sea and land boundary line according to the above step 1-2, the detection information is generated, and the closed area block is divided according to the sea and land boundary line and the circular area line scanned by the radar. Each area is determined by whether the water depth is greater than zero. According to the proportion in the dichotomous statistical area block, if the water depth is greater than zero and the proportion is greater than 50%, it is regarded as a land area and the detection information is 1, otherwise it is regarded as an ocean area and the detection information is 0, and a complete detection information index map is generated. .

Step 2: Traverse all the positions (polar coordinates) of the radar scanning area, convert each position point to latitude and longitude coordinates, and retrieve the detection information:

Step 2-1: Calculate the average distance represented by the unit longitude and latitude, take two points in the horizontal direction within the current display range of the electronic chart, record the longitude difference of the two points, and then call the longitude and latitude-distance of the electronic chart The calculation method is to calculate the distance difference between two points, and then divide the distance difference by the longitude difference to get the average distance represented by the unit longitude. In a similar way, the average distance represented by the unit latitude can also be obtained.

Step 2-2, convert the polar coordinates of each position point (including the position point number and the offset of the azimuth value) into latitude and longitude coordinates. The conversion process is:

The earth is approximated as a sphere, the position point is centered on the radar, and the scanning radius is taken as the radius of the circular area. The polar coordinate information of the position point includes the azimuth value offset and the position point serial number. In step 1, the center latitude and longitude of the radar are set. Scan radius, sampling accuracy, azimuth value, first calculate the upper limit d of the position number sequence number:

dis=c*pno,

pno∈[0,d),

offset∈[0,P),

dis_x=dis*sinang,

dis_y=dis*cosang,

lon=dlon+dis_x/dx,

lat=dlat+dis_y/dy,

Among them, dlon, dlat respectively represent the longitude and latitude of the radar center, dx, dy respectively represent the average distance per unit longitude and the average distance per unit latitude solved in step 2-1, P represents the azimuth value, l represents the scanning radius, and c represents the sampling Accuracy: P, l, c, dlon, dlat and other parameters in step 1 are determined by setting, offset represents the offset of the azimuth value, and pno represents the sequence number of the position point, which belongs to polar coordinate information.

Among them, the latitude-longitude-distance formula adopts the following method: the earth is approximately regarded as a sphere, and the angle between the two points and the center of the earth is calculated according to the latitude and longitude values of the two points, and the angle between the two points and the center of the earth is combined with the angle value of the angle. The radius of the earth, if the radius of the earth is taken as 6371004 meters, the spherical distance between two points can be calculated and regarded as the distance value between the two points.

Step 3: Determine the detection information according to the principle of detection information generation, determine whether it is a land area, and record the area category information of the point.

Determine whether it belongs to a land area. If it is, the result of the type corresponding to the location point is 1, otherwise it is 0. The results of the whole map are divided into regions, and the effect is shown in Figure 2, using thick lines to represent the coastline, separating land areas and non-land areas.

Step 4. Perform an output file operation on all the detection information obtained by the traversal to complete the extraction of the effective detection area.

The detection result set obtained by traversing all position point information, where the result of 1 constitutes the land area of the detection map, and the result of 0 constitutes the water part of the detection map, which is the effective detection area. For all the results obtained in the previous step Values are output to the file to form a detection map. As shown in Figure 3, it is a radar scan effect map of radar enrollment without detection and enrollment files. The land area contains a lot of false information and is an unreal radar target. As shown in Figure 4, after using remote sensing images to identify the land area , The radar automatically avoids the land area when scanning, avoiding the possibility of generating false targets. The original land area in Figure 3 is directly shielded, which improves the accuracy of radar target detection. After using the detection admission file generated in step 4, the land area was filtered to remove a large amount of land clutter, and the effect was good.

The present invention provides a method for extracting effective radar detection areas based on remote sensing images. There are many methods and ways to implement this technical solution. The above are only the preferred embodiments of the present invention. It should be pointed out that for common technologies in this technical field As far as personnel are concerned, without departing from the principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention. All the components that are not clear in this embodiment can be implemented using existing technology.

Claims

A method for extracting a radar effective detection area based on remote sensing images, which is characterized in that it includes the following steps:

Step 1. Set the parameters, use the electronic chart to superimpose the remote sensing image data, and use the water depth information of the electronic chart combined with the edge detection method to generate the detection information index map;

Step 2: Traverse all positions in the radar scanning area, perform polar coordinate conversion on each position point, and obtain the detection information of the position point according to the detection information index map:

Step 3: Judge the detection information, determine whether the location point is a land area, and record the area category information of the location point;

Step 4. Perform the output file operation to complete the extraction of the effective detection area.
The method according to claim 1, wherein step 1 comprises:

Step 1-1, set parameters: set the radar center, scan radius, azimuth value and sampling accuracy, and extract the water depth information from the rectangular area centered on the radar and the side length as the scan diameter based on the water depth information provided by the electronic chart. If the water depth is greater than 0, it is the ocean area, otherwise it is the land area, and the map layer containing the sea and land boundaries is obtained;

Step 1-2: Obtain the remote sensing image of the same scale for the map layer containing the sea and land boundary line, and perform the gray-scale processing on it to obtain the gray-scale image, and then perform edge extraction on the gray-scale image, and the gradient calculation of the edge extraction Sub G[F(i,j)] is:

In the gradient operator G[F(i,j)], F(i,j) represents the gray value of the image, i represents the abscissa of the image, and j represents the ordinate of the image. This operator can reflect the The gray level changes, after the edge detection is completed, the accurate sea and land boundary can be displayed, and then the image size containing the accurate sea and land boundary information is normalized and updated back to the corresponding position;

Step 1-3, according to step 1-2, normalize the size of the image containing the accurate land and sea boundary information and update it back to the corresponding position. Then, all the detection information is generated, which is divided according to the land and sea boundary line and the circular area line scanned by the radar. Enclosed area block, based on whether the water depth is greater than zero and the proportion in the area block is counted. If the water depth is greater than zero and the proportion is greater than 50%, it is determined to be a land area and the detection information is 1; otherwise, it is determined to be an ocean area and the detection information is 0. Complete inspection information index map.
The method according to claim 2, wherein step 2 comprises the following steps:

Step 2-1, calculate the average distance represented by the unit longitude and unit latitude;

Step 2-2, convert the polar coordinates of each location point into latitude and longitude coordinates;

Step 2-3, according to the latitude and longitude coordinates, retrieve the detection information of the corresponding location.
The method according to claim 3, wherein step 2-1 comprises: taking the left middle point and the right middle point in the horizontal direction within the current display range of the electronic chart, and recording the longitude difference of the two points , Then call the longitude, latitude and distance calculation method of the electronic chart to calculate the distance difference between the two points, and then divide the distance difference by the longitude difference to get the average distance represented by the unit longitude; take it within the current display range of the electronic chart In the vertical direction, the top middle point and the bottom middle point are calculated in the same way as the average distance represented by the unit longitude. The average distance represented by the unit latitude is calculated.
The method according to claim 4, wherein step 2-2 comprises:

The earth is approximated as a sphere, the position point is centered on the radar, and the scanning radius is taken as the radius of the circular area. The polar coordinate information of the position point includes the azimuth value offset and the position point serial number. In step 1, the center latitude and longitude of the radar are set. Scan radius, sampling accuracy, azimuth value, first calculate the upper limit d of the position number sequence number:

Then calculate the spherical distance dis between the corresponding position point and the radar center, and the calculation formula is:

dis=c*pno,

pno∈[0,d),

Then obtain the angle information ang from the azimuth offset of the position point:

offset∈[0,P),

According to the trigonometric function conversion, calculate the distance dis_x from the radar center point in the longitude direction that is the horizontal direction and the distance dis_y from the radar center point in the latitude direction that is the vertical direction:

dis_x=dis*sinang,

dis_y=dis*cosang,

According to the average distance represented by the unit longitude and latitude calculated in step 2-1, the longitude value lon and latitude value lat of the corresponding position are calculated, and the formula is as follows:

lon=dlon+dis_x/dx,

lat=dlat+dis_y/dy,

Among them, dlon, dlat respectively represent the longitude and latitude of the radar center, dx, dy respectively represent the average distance per unit longitude and the average distance per unit latitude solved in step 2-1, P represents the azimuth value, l represents the scanning radius, and c represents the sampling Accuracy, the parameters P, l, c, dlon, dlat in step 1 are determined by setting, offset represents the offset of the azimuth value, and pno represents the serial number of the position point, which belongs to polar coordinate information.
The method according to claim 5, wherein step 3 comprises: judging the detection information to determine whether the location point is a land area, if it is, the result of the area category corresponding to the location point is 1, otherwise it is 0.
The method according to claim 6, characterized in that, step 4 comprises: composing the location points whose area category result is 1 into the land area of the detection map, and the location points whose result is 0 compose the water area of the detection map, and the water area of the detection map The part is the effective detection area, and the detection map is displayed in the VTS system.