阅读说明：本技术 一种生成侧扫声呐类正射影像的方法 (Method for generating side-scan sonar orthophoto ) 是由 刘大川 张丽婷 吴少林 严晋 孟涛 孙嘉蔚 杨德鹏 于 2021-08-05 设计创作，主要内容包括：本发明涉及一种生成侧扫声呐类正射影像的方法,该方法基于横向分辨率约束和纵向影像综合处理相结合,其包括如下步骤：S1、影像采集：对侧扫声呐接收到的信号进行采集并生成非正射影像；S2、计算约束因子；S3、计算位移差；S4、像素综合处理：对S1中的非正射影像进行综合处理,得到预处理影像；S5、影像增强处理：对S4得到的预处理影像进行影像增强处理,得到类正射影像,类正射影像沿航线方向的分辨率与垂直于航线方向的分辨率相匹配。本发明的方法可以让侧扫声呐采集到的影像畸变大幅度降低,提高了量取性,降低了测量误差。(The invention relates to a method for generating side-scan sonar orthophoto images, which is based on the combination of transverse resolution constraint and longitudinal image comprehensive processing and comprises the following steps: s1, image acquisition: collecting signals received by the side scan sonar and generating a non-orthographic image; s2, calculating a constraint factor; s3, calculating a displacement difference; s4, pixel integration: comprehensively processing the non-orthoimage in the S1 to obtain a preprocessed image; s5, image enhancement: and performing image enhancement on the preprocessed image obtained in the step S4 to obtain a quasi-orthoimage, wherein the resolution of the quasi-orthoimage along the flight line direction is matched with the resolution perpendicular to the flight line direction. The method can greatly reduce the image distortion collected by the side scan sonar, improve the measuring performance and reduce the measuring error.)

1. A method for generating side-scan sonar orthophoto images is characterized in that: the method uses a mode of combining the transverse resolution constraint and the longitudinal image comprehensive processing, and comprises the following steps:

s1, image acquisition: collecting signals received by the side scan sonar and generating a non-orthoscopic image, wherein the non-orthoscopic image comprises non-orthoscopic transverse pixels and non-orthoscopic longitudinal pixels, and performing slant distance correction on the non-orthoscopic image;

s2, calculating a constraint factor: when side scan sonar is used for measurement, the known scan width L and the single-row pixel number n of the non-orthoimage vertical to the flight path are calculated, and the resolution ratio of the non-orthoimage vertical to the flight path is calculatedThe obtained r is a constraint factor;

s3, calculating a displacement difference: the signals received by the side scan sonar in S1 carry position information, the position information comprises a transverse coordinate x and a longitudinal coordinate y, a plurality of position information can be obtained by sampling along with the movement of the air route and the fixed sampling rate of the side scan sonar, and the displacement difference between two adjacent groups of position information is calculated，

Wherein the content of the first and second substances,is the difference between the lateral coordinates of two adjacent sets of position information,the difference between the longitudinal coordinates of two adjacent sets of position information;

s4, pixel integration: comprehensively processing the non-orthoimage in the S1 to obtain a preprocessed image;

s5, image enhancement: and performing image enhancement processing on the preprocessed image obtained in the step S4 to obtain a quasi-orthoimage, wherein the resolution of the quasi-orthoimage along the flight line direction is matched with the resolution perpendicular to the flight line direction.

2. The method of generating a side-scan sonar-like ortho image according to claim 1, comprising: in the step S4, the non-orthonormal vertical pixels in the non-orthonormal image obtained in step S1 are adjusted by using a contrast method according to the following mathematical formula:

wherein N is the theoretical pixel number along the flight path direction in the standard proportion image, D is the displacement difference obtained in S3, and r is the constraint factor obtained in S2;

in addition, it is known that the number of scans between the initial position point and the final position point along the course direction in the non-orthographic image is Y, and Y is compared with N, and there are cases where:

when Y isAt time N, the number of pixels of the non-orthometric vertical pixels in S1 is complemented;

when Y isN, the non-orthonormal vertical pixels in S1 are retained;

when Y isN, the number of pixels of the non-orthonormal vertical pixels in S1 is reduced by a screen.

3. The method of generating a side-scan sonar-like ortho image according to claim 2, comprising: the compensation of the number of pixels of the non-orthometric vertical pixels in S1 adopts a newton interpolation formula method.

4. The method of generating a side-scan sonar-like ortho image according to claim 2, comprising: the filtering for the number of pixels of the non-orthonormal vertical pixels in S1 adopts a moving average processing method.

5. The method of generating a side-scan sonar-like ortho image according to claim 1, comprising: in S5, the image enhancement processing is performed by a median filter processing method.

Technical Field

The invention relates to the technical field of side-scan sonars, in particular to a method for generating side-scan sonar orthophotos.

Background

Side scan sonar is equipment for detecting submarine landform and underwater objects by using echo sounding principle, and is also called side sonar or submarine landform instrument.

The utility model with the publication number of CN205210305U discloses a side scan sonar device, which comprises a sonar device and a power supply device; the sonar device is arranged in the protective shell; the power supply device is provided with a fully-closed shell, a rechargeable battery and a battery management circuit board are arranged in the fully-closed shell, the battery management circuit board controls the charging or discharging of the rechargeable battery to be used by the sonar device, a magnetic switch is arranged in the fully-closed shell, and a magnetic switch hole is formed in the outer surface of the fully-closed shell; the magnet is detachably arranged in the magnetic switch hole; the magnet controls the opening and closing of the magnetic switch; the magnetic switch controls the connection and disconnection of an output circuit of the power supply device; the power supply device is electrically connected with the sonar device through a lead; the wires are connected with the power supply device and the sonar device in a pluggable and pluggable manner.

The lateral resolution of the side-scan sonar image obtained by the side-scan sonar device mainly relates to the sampling frequency and the scan width, and the scan width of the side-scan sonar is generally known, while the sampling rate of the side-scan sonar is generally unchanged; the longitudinal resolution of the image is mainly related to the measuring speed of the carrier and the working frequency of the sonar. Because the side scan sonar can receive the influence of measuring speed when measuring, so there is geometric distortion in the collection image that the side scan sonar generated, especially along the inconsistent with the image resolution ratio of perpendicular to course along the course for there is great difference in collection image and the actual standard proportion image, leads to the target object size on the image to have great deviation, and then leads to the measurability reduction of image, and data measurement error grow.

Disclosure of Invention

Aiming at the defects in the prior art, the invention aims to provide a method for generating side-scan sonar orthophoto images.

The above object of the present invention is achieved by the following technical solutions:

a method for generating side-scan sonar orthophoto images is based on the combination of transverse resolution constraint and longitudinal image comprehensive processing, and comprises the following steps:

s1, image acquisition: collecting signals received by the side scan sonar and generating a non-orthoscopic image, wherein the non-orthoscopic image comprises non-orthoscopic transverse pixels and non-orthoscopic longitudinal pixels, and performing skew correction on the non-orthoscopic image;

s2, calculating a constraint factor: when side scan sonar is used for measurement, the known scan width L and the single-row pixel number n of the non-orthoimage vertical to the flight path are calculated, and the resolution ratio of the non-orthoimage vertical to the flight path is calculatedThe obtained r is a constraint factor;

s3, calculating a displacement difference: the signals received by the side scan sonar in S1 carry position information, the position information comprises a transverse coordinate x and a longitudinal coordinate y, a plurality of position information can be obtained by sampling along with the movement of the air route and the fixed sampling rate of the side scan sonar, and the displacement difference between two adjacent groups of position information is calculated，

Wherein the content of the first and second substances,is the difference between the lateral coordinates of two adjacent sets of position information,the difference between the longitudinal coordinates of two adjacent sets of position information;

s4, pixel integration: comprehensively processing the non-orthoimage in the S1 to obtain a preprocessed image;

s5, image enhancement: and performing image enhancement processing on the preprocessed image obtained in the step S4 to obtain a quasi-orthoimage, wherein the resolution of the quasi-orthoimage along the flight line direction is matched with the resolution perpendicular to the flight line direction.

By adopting the technical scheme, firstly, image acquisition is carried out, and a non-orthoscopic image with distortion is generated, wherein the non-orthoscopic image comprises non-orthoscopic transverse pixels and non-orthoscopic longitudinal pixels, a transverse image formed by the non-orthoscopic transverse pixels can correctly reflect the transverse size of a target object, but a longitudinal image formed by the non-orthoscopic longitudinal pixels cannot reflect the longitudinal size of the target object due to distortion. Therefore, the constraint factor r and the displacement difference D are calculated, and then the non-orthoimage can be comprehensively processed through the constraint factor r and the displacement difference D to obtain a preprocessed image, wherein the preprocessed image can correctly reflect the size of the target object. And finally, enhancing the preprocessed image to obtain a final similar orthoimage, wherein the similar orthoimage has no size change compared with the preprocessed image, but the definition is greatly improved, and the image is clearer. The measurement performance of the image acquired by the side scan sonar is improved through the steps, and the measurement error is reduced.

The present invention in a preferred example may be further configured to: in the S4 process, the non-orthoscopic longitudinal image in the non-orthoscopic images obtained in S1 is adjusted, and the adjustment is performed by using a contrast method according to the mathematical formula:

wherein N is the theoretical pixel number along the flight path direction in the standard proportion image, D is the displacement difference obtained in S3, and r is the constraint factor obtained in S2;

in addition, it is known that the number of scans between the initial position point and the final position point along the course direction in the non-orthographic image is Y, and Y is compared with N, and there are cases where:

when Y isAt time N, the number of pixels of the non-orthometric vertical pixels in S1 is complemented;

when Y isN, the non-orthonormal vertical pixels in S1 are retained;

when Y isN, the number of pixels of the non-orthonormal vertical pixels in S1 is reduced by a screen.

By adopting the technical scheme, the theoretical pixel number N is obtained through calculation according to the constraint factor r and the displacement difference D, and the theoretical pixel number N is compared with the scanning times Y of the side scan sonar. If it is YN, the vertical compression of the non-orthoimage is explained, so that the height of the non-orthoimage is too low, and the number of pixels in the non-orthoimage vertical direction needs to be complemented; if it is YN, the longitudinal proportion of the non-orthoimage is proper and does not need to be adjusted; if it is YN means that the non-orthonormal image is stretched in the vertical direction, and the height of the non-orthonormal image is too high, so that the number of pixels in the non-orthonormal vertical direction needs to be reduced.

The present invention in a preferred example may be further configured to: the pixel value of the non-orthoscopic longitudinal image in the step S1 is complemented by using a newton interpolation formula method.

By adopting the technical scheme, the Newton interpolation formula has the advantages of smaller calculation amount and capability of improving the operation efficiency, so that the preprocessed image can be quickly obtained.

The present invention in a preferred example may be further configured to: the filtering of the pixel values of the non-orthoimage in S1 adopts a moving average processing method.

By adopting the technical scheme, the moving average processing method has the advantages of simple algorithm, small calculation amount, capability of improving the operation efficiency and convenience and rapidness in obtaining the preprocessed image.

The present invention in a preferred example may be further configured to: in S5, the image enhancement processing is performed by a median filter processing method.

By adopting the technical scheme, the advantage of the median filtering processing is that the processing mode of the median filtering is a nonlinear method, the method is very effective in the aspect of eliminating image noise points, meanwhile, the sharp edges of the image can be protected, and proper points are selected to replace the values of the pollution points, so the processing effect is good.

In summary, the invention includes at least one of the following beneficial technical effects:

1. the quasi-ortho image is obtained by adjusting and processing the non-ortho image, the size of a target object can be correctly reflected, the measuring capability of the side scan sonar image is ensured, and the data measurement error is reduced;

2. the calculation formulas of the constraint factor and the displacement difference are simple and convenient, the calculation amount is reduced, the image adjusting speed is improved, and the working efficiency is further improved;

3. in the adjusting process of the non-orthometric longitudinal pixels, the pixels are adjusted by a Newton interpolation formula method and a moving average method, so that the calculation is simple and convenient, and the image generation speed is further improved.

Drawings

FIG. 1 is an overall step diagram of the method of the present invention.

Fig. 2 is a non-orthoimage obtained after S1.

FIG. 3 is a schematic view of the pixel layout of an image formed after the scan width and the number of scans are set during the operation of a side scan sonar;

fig. 4 is a preprocessed image obtained after S4.

Fig. 5 is an ortho-like image obtained after S5.

Fig. 6 is a logic diagram of the comparison method in S3.

In the figure, 1, a target object.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings.

The invention discloses a method for generating side-scan sonar orthophoto images, which is based on the combination of transverse resolution constraint and longitudinal image comprehensive processing. The starting point of the combination is that the image distortion generated by the side-scan sonar is basically caused by the speed of navigation, and the sampling rate of the side-scan sonar is basically fixed, so the transmitting and receiving frequency of the signals of the side-scan sonar is fixed no matter what the navigation speed is, and therefore, when the navigation speed is high, the image is compressed, and the longitudinal height is shortened; when the navigation speed is slow, the image is stretched and the longitudinal height is elongated. The scan width of the side scan sonar is also known, and determines the image frame width, and the scan width is constant, so the image frame width is also constant. Thus, the width of the images is communicated, and the height of the images is changed due to the speed of navigation.

Referring to fig. 1, an overall step diagram of the method specifically includes the following steps:

s1, image acquisition: referring to fig. 2, a side scan sonar transmits and receives signals, and generates a non-orthoimage, which is an image with distortion, by collecting the received signals. The non-orthoscopic image comprises non-orthoscopic transverse pixels and non-orthoscopic longitudinal pixels, wherein the transverse image generated by the non-orthoscopic transverse pixels is basically not distorted and can reflect the transverse size of the target object 1, but the longitudinal image generated by the non-orthoscopic longitudinal pixels is severely distorted and cannot reflect the longitudinal size of the target object 1. In the image acquisition process, the image of the non-orthoimage needs to be subjected to slant range correction, so that the accuracy of the size of the image of the non-orthoimage is further improved, and a foundation is laid for the next correction;

s2, calculating a constraint factor: the constraint factor is used to constrain the non-orthonormal vertical pixels. Referring to fig. 3, the side scan sonar has been set for the scan width at the time of measurement, and the scan width is denoted by L. It is also known that the number of pixels of a single column of the non-orthonormal image in the direction perpendicular to the course, i.e. all the pixels of a single column of non-orthonormal longitudinal pixels, is denoted by n, and the swath width L and the number of pixels of a single column n are not changed with the change of the navigation speed.

Referring to FIG. 3, the non-orthonormal image resolution perpendicular to the course direction, which is represented by r, can be calculated by the swath width L and the number of pixels n of the non-orthonormal image in a single row perpendicular to the course directionAnd r obtained by calculation is not only the resolution perpendicular to the flight path direction, but also a constraint factor.

S3, calculating a displacement difference: the displacement difference is the amount of change in position between two samplings of the side scan sonar. Referring to fig. 3, the signal received by the side scan sonar in S1 carries the position information of the target object 1, and this position information is displayed in the form of two-dimensional coordinates including lateral coordinates and longitudinal coordinates, the lateral coordinates being represented by x and the longitudinal coordinates being represented by y. During the process that the side scan sonar moves along with the flight path, a plurality of position information, namely a plurality of transverse coordinates x and a plurality of longitudinal coordinates y, can be collected through the fixed sampling rate of the side scan sonar. For example, fig. 3 is an image of the target object 1 during the sampling process, and it should be noted that the shape of the target object 1 in the figure is only an illustration, and the shape of the actual target object 1 is much more complicated. WhereinAndthat is, two adjacent sets of position information on the contour of the target object 1 in the sampling, so that the position difference between the two adjacent sets of position information can be calculated, the displacement difference is represented by D, and the calculation formula isWherein, in the step (A),as a difference between the lateral coordinates of two adjacent sets of position information, i.e.，As a difference between the longitudinal coordinates of two adjacent sets of position information, i.e.；

S4, pixel integration: referring to fig. 4, the process of pixel synthesis processing may perform synthesis processing on the non-orthoimage obtained in S1, and obtain a preprocessed image after the synthesis processing;

s5, image enhancement: referring to fig. 5, the image enhancement processing may perform image enhancement on the preprocessed image obtained in S4, mainly enhance the definition, and may display the target object 1 more intuitively, which is convenient for measurement, and obtain an ortho-like image through the image enhancement processing, where the resolution of the ortho-like image in the eye-route direction is adapted to the resolution perpendicular to the route direction, so as to correctly represent the transverse size and the longitudinal size of the target object 1.

The above is the main step of the whole method, and the details thereof are discussed in detail below:

referring to fig. 2 to 4, the pixel integration processing mentioned in S4 is performed by adjusting the number of pixels of the non-orthometric vertical pixels in the non-orthometric image in S1 to match the number of pixels of the non-orthometric vertical pixels with the number of pixels of the non-orthometric horizontal image, so that the reason for the adjustment is that the horizontal image composed of the non-orthometric horizontal pixels is not distorted, and the distortion degree of the vertical image generated by the non-orthometric vertical pixels can be greatly reduced after the adjustment.

Referring to FIG. 6, the adjustment is performed by using a comparison method based on a mathematical formulaWherein N is the theoretical pixel number along the flight path direction in the standard proportion image, D is the displacement difference obtained in S3, and r is the constraint factor obtained in S2;

in addition, because the sampling rate of the side scan sonar is fixed, the scanning times between the initial position point and the end position point along the course direction in the non-orthographic image can be determined, and the scanning times is represented by Y. Next, comparing Y with N, there are three cases as follows:

1) when Y isN is the number of pixels in the non-orthonormal vertical direction in the non-orthonormal video in S1. Because of YN means that the number of non-orthonormal vertical pixels is insufficient, which is generally caused by over-speeding, and this results in a shorter vertical height of the non-orthonormal image, so that the number of pixels needs to be supplemented. The method for complementing the pixel number adopts a Newton interpolation formula method, the application of the Newton interpolation formula method needs to use a Newton interpolation formula, and the Newton interpolation formula is the prior art and is not described in more detail here. The advantage of using Newton interpolation for complement is that the calculation amount is small, and the operation efficiency can be improved;

2) when Y isN, the non-orthonormal vertical pixels in S1 are retained. Because of YN means that the number of pixels of the non-orthometric vertical pixels in the non-orthometric image matches the theoretical number of pixels, and the vertical height of the non-orthometric image is not distorted at this time, and the size can be correctly reflected, so that adjustment is not required. It should be noted that the so-called YN is not necessarily considered to be completely identical to N, and is preferably approximately equal to Y, and the range is controlled to be within 5%, that is, within 5%All of them can be regarded as YN；

3) When Y isAt time N, the number of pixels of the non-orthonormal vertical pixels in the non-orthonormal image in S1 is reduced by a screen. Because of YN means that the number of pixels in the non-orthoscopic image is too large, which is generally caused by too slow navigation speed, and thus the vertical height of the non-orthoscopic image is too long, so that the number of pixels needs to be reduced. The method for filtering the pixel number adopts a moving average method, which is the prior art and is not described in more detail herein. The advantage of using the sliding average method for screening and reducing is that the algorithm is simple and convenient, the calculated amount is small, and the operation efficiency can be improved.

Referring to fig. 5, the image enhancement processing in S5 is to substantially eliminate the noise in the pre-processed image obtained in S4, so as to obtain a clearer quasi-orthoimage. The mode used for enhancing the image in this process is a median filtering mode, and the image processing method of median filtering is the prior art and will not be described in great detail here. The reason for adopting the median filtering is that the median filtering is a nonlinear method, which is very effective in the aspect of eliminating image noise points, and meanwhile, the method can protect sharp edges of images, and selects proper points to replace the values of pollution points, so the processing effect is good.

Referring to fig. 1 to 5, through the above steps, in S1, the side scan sonar transmits and receives signals to generate a non-orthoimage, which is a distorted image; through S2, calculating a constraint factor, wherein the constraint factor is used for constraining the non-orthoimage; through S3, calculating a displacement difference, wherein the displacement difference is used for verifying whether the navigation speed of the side scan sonar meets the requirement; through S4, the non-orthoimage is adjusted to be a preprocessed image, and the longitudinal height of the preprocessed image is matched with the actual size of the target object 1; through S5, the preprocessed image is enhanced into an ortho-like image, and the sharpness of the ortho-like image is improved compared to the preprocessed image.

The embodiments of the present invention are preferred embodiments of the present invention, and the scope of the present invention is not limited by these embodiments, so: all equivalent changes made according to the structure, shape and principle of the invention are covered by the protection scope of the invention.