CN108537853B - Underwater sonar image compression transmission method - Google Patents

Underwater sonar image compression transmission method Download PDF

Info

Publication number
CN108537853B
CN108537853B CN201810193720.2A CN201810193720A CN108537853B CN 108537853 B CN108537853 B CN 108537853B CN 201810193720 A CN201810193720 A CN 201810193720A CN 108537853 B CN108537853 B CN 108537853B
Authority
CN
China
Prior art keywords
sparse
wavelet transform
image
sonar image
compressed sensing
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.)
Expired - Fee Related
Application number
CN201810193720.2A
Other languages
Chinese (zh)
Other versions
CN108537853A (en
Inventor
付晓梅
张静
常帅
张亮
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Priority to CN201810193720.2A priority Critical patent/CN108537853B/en
Publication of CN108537853A publication Critical patent/CN108537853A/en
Application granted granted Critical
Publication of CN108537853B publication Critical patent/CN108537853B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T9/00—Image coding
    • G06T9/007—Transform coding, e.g. discrete cosine transform

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Discrete Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

The invention discloses a compression transmission method of an underwater sonar image, which is based on the compressed sensing of sparse wavelet transform and comprises the following steps: 1) sparse representation is carried out on the sonar image through sparse wavelet transform, so that original image data are changed into sparse signals which can be used for compressed sensing; 2) the sparse signals are projected into the measurement matrix, so that the sparse signals are converted from high dimensionality to low dimensionality, and the image is compressed; 3) reconstructing the sparse signal generated in the step 1) by adopting a compressed sensing reconstruction algorithm, namely recovering high-dimensional data by using low-dimensional data; 4) the reconstructed sparse signal is subjected to the inverse transform of the sparse wavelet transform, that is, the inverse transform of step 1) is performed, thereby restoring the original sonar image. The invention can represent the original image by less data and realize higher compression ratio.

Description

Underwater sonar image compression transmission method
Technical Field
The invention relates to the field of underwater sonar image transmission, in particular to a compression transmission method for underwater sonar images.
Background
Sonar plays a very important role in underwater exploration. Sonar images of the underwater environment may be transmitted over an underwater acoustic channel. However, the characteristics of the underwater acoustic channel are as follows: doppler dispersion, multipath effects, and in particular limited bandwidth, make transmitting images over underwater acoustic channels a challenging task.
In order to save the cost of data storage and transmission, image compression is widely applied to sonar image processing. Discrete cosine transform is used in real-time image processing, but this method is prone to blocking effects; discrete wavelet transform can achieve compression of an image by decomposing signals on different scales, but edges and contours of a reconstructed image become blurred; a compressed sensing-based Bandelets transformation is a typical multi-scale geometric analysis image representation method, and can adaptively track the geometric regular direction of a sonar image, but under the condition that the image is known, the method cannot meet the requirement of underwater real-time transmission.
Disclosure of Invention
The invention provides a compression transmission method of an underwater sonar image, which is based on the compressed sensing of sparse wavelet transform and applied to the transmission of the underwater sonar image in an underwater acoustic channel, and is described in detail as follows: a compression transmission method of an underwater sonar image is based on compressed sensing of sparse wavelet transform and comprises the following steps:
1) sparse representation is carried out on the sonar image through sparse wavelet transform, so that original image data are changed into sparse signals which can be used for compressed sensing;
2) The sparse signals are projected into the measurement matrix, so that the sparse signals are converted from high dimensionality to low dimensionality, and the image is compressed;
3) reconstructing the sparse signal generated in the step 1) by adopting a compressed sensing reconstruction algorithm, namely recovering high-dimensional data by using low-dimensional data;
4) the reconstructed sparse signal is subjected to the inverse transform of the sparse wavelet transform, that is, the inverse transform of step 1) is performed, thereby restoring the original sonar image.
Wherein the sparse wavelet transform and the inverse transform of the sparse wavelet transform are implemented based on a sparse wavelet transform basis.
The forming process of the sparse wavelet transform basis specifically comprises the following steps:
the wavelet transform matrix is
Figure BDA0001592451300000021
First, a new wavelet transmission matrix is generated
Figure BDA0001592451300000022
Then W isrExpanding the transformation coefficient of each row by m times, regarding the m rows of coefficients as a group, dividing the coefficient of each row in the group by the first row, shifting a time domain sampling point by delta relative to the previous row, and filling zero in other positions;
the new matrix formed after transformation is a transformation base with sparse characteristic, called sparse wavelet transformation base, and expressed as
Figure BDA0001592451300000023
Wherein N ═ mr, K ═ r + (m-1) Δ.
Further, the measurement matrix is a gaussian measurement matrix.
The compressed sensing reconstruction algorithm is a basis pursuit reconstruction algorithm.
The technical scheme provided by the invention has the beneficial effects that:
1. since sparse wavelet transform can realize signal representation more sparsely than wavelet transform, the invention is used for processing underwater sonar images by using a compressed sensing technology, can represent original images by less data and realizes higher compression ratio;
2. the method can transmit sonar images with less data volume and improve the bandwidth utilization rate, and compared with wavelet transformation, the method realizes better peak signal-to-noise ratio and structural similarity index with higher bandwidth utilization rate under the condition of the same sampling rate;
3. compared with the traditional wavelet transform, the method improves the quality of the restored image at the receiving end.
Drawings
FIG. 1 is a flow chart of a method for compressing and transmitting underwater sonar images;
FIG. 2 is a graph of peak signal-to-noise ratio versus compression ratio;
FIG. 3 is a graph of structural similarity versus compression ratio.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention are described in further detail below.
Compressed sensing is an image compression method that has been widely used in recent years because of the ability to efficiently transmit images.
The embodiment of the invention provides a compression transmission method of an underwater sonar image, which uses a compression sensing technology for sonar image compression to solve the problem that excessive bandwidth is needed for sonar image transmission underwater and improve the bandwidth utilization rate. In the method, sparse wavelet transform is used for sparse representation of compressed sensing, the original image is represented by the least data, and higher peak signal-to-noise ratio and higher structural similarity are realized at a receiving end.
The sparse representation is an important step in compressed sensing, and the more sparse representation can usually realize higher compression ratio, that is, a sonar image is represented by less data, so that a certain bandwidth is saved.
The embodiment of the invention adopts sparse wavelet transform to realize sparse representation, and compared with wavelet transform, the method can use less data to represent the image, realize more sparse representation and ensure the quality of the sonar image.
Example 1
A compression transmission method of underwater sonar images is disclosed, referring to fig. 1, and comprises the following steps:
101: sparse representation is carried out on the sonar image through sparse wavelet transform, so that original image data are changed into sparse signals which can be used for compressed sensing;
102: the sparse signal is projected into a measurement matrix phi, so that the sparse signal is converted from high dimension (1 multiplied by P) to low dimension (1 multiplied by Q), the image is compressed, and the compression ratio is
Figure BDA0001592451300000031
Where the measurement matrix Φ needs to satisfy finite equidistant properties, such as: gauss measurement matrix phiGIs a commonly used measurement matrix, can meet the finite equidistant property with a larger probability, and realizes the compression of sparse signals, phiGIs a random matrix of standard normal distribution with a mean of 0 and a variance of 1.
The embodiment of the invention is to measure the matrix phi by GaussGFor example, the measurement matrix Φ is illustrated, and when the measurement matrix Φ is specifically implemented, the measurement matrix Φ may also be a measurement matrix in another form.
103: a compressed sensing reconstruction algorithm is adopted, namely, high-dimensional data are recovered by using low-dimensional data, and the sparse signal generated in the step 101 is reconstructed;
the compressed sensing reconstruction algorithm may adopt a basis pursuit reconstruction algorithm known in the art, and may be other reconstruction algorithms when being specifically implemented, which is not limited in the embodiment of the present invention.
104: the reconstructed sparse signal is subjected to the inverse transform of the sparse wavelet transform, that is, the inverse transform of step 101, to restore the original sonar image.
In summary, in the embodiment of the present invention, the compressed sensing technology is used for sonar image compression through the above steps 101 to 104, so that the original image can be represented by less data, and a higher compression ratio can be realized.
Example 2
The scheme of example 1 is further described below with reference to a specific calculation formula and fig. 1, and is described in detail below:
201: a sonar image processing process based on sparse wavelet transform;
referring to fig. 1, firstly, selecting a sonar image, and performing sparse representation on the original sonar image through sparse wavelet transform; sampling the sparse signal by a compressed sensing technology; reconstructing the sparse signal using a reconstruction algorithm; and recovering the original sonar image through the inverse transformation of the sparse wavelet transform.
202: the generation principle of a sparse wavelet transform basis;
in step 101 and step 104, the inverse sparse wavelet transform and the inverse sparse wavelet transform are performed, respectively. And performing the pair of transforms is based on a sparse wavelet transform basis. The sparse wavelet transform basis is formed based on the wavelet transform basis, and the forming process is mainly divided into three steps.
Suppose that the wavelet transform matrix is
Figure BDA0001592451300000041
First, a new wavelet transmission matrix is generated
Figure BDA0001592451300000042
Then W isrIs expanded by m times, the m lines of coefficients are regarded as a group, and each line of coefficients (except the first line) in the group is subjected to time domain sampling relative to the previous line The samples are shifted by Δ and other positions are zero-padded.
The new matrix formed after transformation is a transformation base with sparse characteristic, called sparse wavelet transformation base, and can be expressed as
Figure BDA0001592451300000043
Wherein N ═ mr, K ═ r + (m-1) Δ,
Figure BDA0001592451300000044
is a real number field (
Figure BDA0001592451300000045
Representing a matrix of dimensions K x K,
Figure BDA0001592451300000046
a matrix of dimension r x r is represented,
Figure BDA0001592451300000047
representing a matrix with dimension N × K), the value of K is a positive integer.
203: and evaluating the transmission performance of the sonar image.
Wherein, the peak signal-to-noise ratio is used for evaluating an objective standard of the image, and the calculation formula is as follows:
Figure BDA0001592451300000048
wherein I represents the pixel matrix of the original image,
Figure BDA0001592451300000049
a matrix of pixels representing the image recovered at the receiving end,
Figure BDA00015924513000000410
as a pixel matrix I and
Figure BDA00015924513000000411
the calculation formula is as follows:
Figure BDA00015924513000000412
in the above formula n is I (or
Figure BDA00015924513000000413
) I and j are position indexes of the pixel matrix.
In summary, the embodiments of the present invention implement sparse representation by using sparse wavelet transform, and compared with wavelet transform, the method can represent an image with less data, implement more sparse representation, and ensure the quality of a sonar image.
Example 3
The feasibility of the protocols of examples 1 and 2 is verified below with reference to FIGS. 2 and 3, and the specific experimental data, as described in detail below:
In the embodiment of the invention, sonar images, namely, a training user and an airplan Windscreen Detaill observed by an ARIS Explorer 3000 sonar system are selected, pixels are 256 multiplied by 256, sparse representation based on wavelet transformation and sparse wavelet transformation is respectively carried out on an original image, a comparison result of a peak signal-to-noise ratio is shown in figure 2, and a comparison result of structural similarity is shown in figure 3.
As can be seen from fig. 2, two different sonar images respectively have different peak signal-to-noise ratios after being subjected to sparse wavelet transform-based compression reconstruction and wavelet transform-based compression reconstruction. Under the same compression ratio, the compression reconstruction based on the sparse wavelet transform has higher peak signal-to-noise ratio which is improved by about 5dB compared with the compression reconstruction based on the wavelet transform.
As can be seen from fig. 3, under the same compression ratio, the compression reconstruction based on the sparse wavelet transform has higher structural similarity, i.e. better meets the requirements of human visual nerves and perception on image quality. Meanwhile, it can be seen that different sonar images have greatly different structural similarity after compression and reconstruction, which is related to the structural characteristics of the images.
Those skilled in the art will appreciate that the drawings are only schematic illustrations of preferred embodiments, and the above-described embodiments of the present invention are merely provided for description and do not represent the merits of the embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (3)

1. A compression transmission method of an underwater sonar image is characterized in that the compression transmission method is based on compressed sensing of sparse wavelet transform, and comprises the following steps:
1) sparse representation is carried out on the sonar image through sparse wavelet transform, so that original image data are changed into sparse signals which can be used for compressed sensing;
2) the sparse signals are projected into the measurement matrix, so that the sparse signals are converted from high dimensionality to low dimensionality, and the image is compressed;
3) reconstructing the sparse signal generated in the step 1) by adopting a compressed sensing reconstruction algorithm, namely recovering high-dimensional data by using low-dimensional data;
4) performing inverse transformation of sparse wavelet transform on the reconstructed sparse signals, namely performing the inverse transformation of the step 1), thereby recovering the original sonar image;
wherein the sparse wavelet transform and the inverse transform of the sparse wavelet transform are implemented based on a sparse wavelet transform basis;
The forming process of the sparse wavelet transform basis specifically comprises the following steps:
the wavelet transform matrix is
Figure FDA0003189757300000011
First, a new wavelet transmission matrix is generated
Figure FDA0003189757300000012
r<K, then WrIs expanded by a factor of m, the m lines of coefficients are treated as a group, and each line of coefficients in the group is divided byThe first row is used for shifting a time domain sampling point by delta relative to the previous row, and other positions are filled with zero;
the new matrix formed after transformation is a transformation base with sparse characteristic, called sparse wavelet transformation base, and expressed as
Figure FDA0003189757300000013
Wherein N ═ mr, K ═ r + (m-1) Δ.
2. The method for compressing and transmitting the underwater sonar image according to claim 1, wherein the measurement matrix is a gaussian measurement matrix.
3. The method for compressing and transmitting the underwater sonar image according to claim 1, wherein the compressed sensing reconstruction algorithm is a basis pursuit reconstruction algorithm.
CN201810193720.2A 2018-03-09 2018-03-09 Underwater sonar image compression transmission method Expired - Fee Related CN108537853B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810193720.2A CN108537853B (en) 2018-03-09 2018-03-09 Underwater sonar image compression transmission method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810193720.2A CN108537853B (en) 2018-03-09 2018-03-09 Underwater sonar image compression transmission method

Publications (2)

Publication Number Publication Date
CN108537853A CN108537853A (en) 2018-09-14
CN108537853B true CN108537853B (en) 2021-10-12

Family

ID=63485639

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810193720.2A Expired - Fee Related CN108537853B (en) 2018-03-09 2018-03-09 Underwater sonar image compression transmission method

Country Status (1)

Country Link
CN (1) CN108537853B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110708561A (en) * 2019-09-12 2020-01-17 北京理工大学 Underwater information acquisition and transmission method based on compressed sensing and channel coding
CN112114300B (en) * 2020-09-14 2022-06-21 哈尔滨工程大学 Underwater weak target detection method based on image sparse representation
CN112882115B (en) * 2021-04-30 2021-07-27 湖南师范大学 Magnetotelluric signal denoising method and system based on GWO optimized wavelet threshold

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103854250A (en) * 2014-03-17 2014-06-11 广东石油化工学院 Self-synchronizing image watermarking method applied to wireless multimedia sensor network
CN105894547A (en) * 2016-05-06 2016-08-24 南昌航空大学 Image processing method based on group-wave transformation compressed sensing
CN107016656A (en) * 2017-04-01 2017-08-04 中国科学院光电技术研究所 Wavelet Sparse Basis Optimization Method in Image Reconstruction Based on Compressed Sensing

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7271747B2 (en) * 2005-05-10 2007-09-18 Rice University Method and apparatus for distributed compressed sensing
US8014616B2 (en) * 2007-11-02 2011-09-06 Siemens Aktiengesellschaft System and method for fixed point continuation for total variation based compressed sensing imaging

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103854250A (en) * 2014-03-17 2014-06-11 广东石油化工学院 Self-synchronizing image watermarking method applied to wireless multimedia sensor network
CN105894547A (en) * 2016-05-06 2016-08-24 南昌航空大学 Image processing method based on group-wave transformation compressed sensing
CN107016656A (en) * 2017-04-01 2017-08-04 中国科学院光电技术研究所 Wavelet Sparse Basis Optimization Method in Image Reconstruction Based on Compressed Sensing

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
"Application of Bayesian Compressive Sensing in IRUWB Channel Estimation";Song Liu.etc;《COMMUNICATION THEORIES & SYSTEMS》;20170531;全文 *
"压缩感知及其图像处理应用研究进展与展望";任越美等;《自动化学报》;20140831;第40卷(第8期);全文 *

Also Published As

Publication number Publication date
CN108537853A (en) 2018-09-14

Similar Documents

Publication Publication Date Title
CN108537853B (en) Underwater sonar image compression transmission method
Hemalatha et al. Wavelet transform based steganography technique to hide audio signals in image
CN107016656B (en) Wavelet sparse basis optimization method in image reconstruction based on compressed sensing
CN107392952B (en) no-reference mixed distortion image quality evaluation method
CN109309840A (en) Compressed sensing image encoding method based on noise shaping, storage medium
CN108419083B (en) Image multilevel wavelet full subband compressed sensing coding method
Mishra et al. Comparative analysis of wavelet basis functions for ECG signal compression through compressive sensing
Kekre et al. ImageCompression Using Real Fourier Transform, Its Wavelet Transform And Hybrid Wavelet With DCT
Bi et al. Image compressed sensing based on wavelet transform in contourlet domain
CN113077376A (en) Color image zero-watermarking algorithm based on ternary decimal order extremely complex exponential transformation
Alsayyh et al. A Novel Fused Image Compression Technique Using DFT, DWT, and DCT.
CN107612555B (en) Improved sparsity self-adaptive matching tracking algorithm based on dichotomy
CN107948644B (en) Underwater image compression method and transmission method
Rostam et al. Medical and color images compression using new wavelet transformation
Chinnarao et al. Improved image de noising algorithm using dual tree complex wavelet transform
Reddy et al. 2D dual-tree complex wavelet transform based image analysis
Santoso et al. Color image compression using orthogonal wavelet viewed from decomposition level and peak signal to noise ratio
CN105184832B (en) A method of improving the image reconstruction of Noise Variance Estimation
CN102833537A (en) Compressed sensing image sparse method based on Contourlet transformation
CN113658282A (en) Image compression and decompression method and device
Pandian et al. Effects of decomposition levels of wavelets in image compression algorithms
Bo et al. Image compression based on wavelet transform
Nahar A compression original image based on the DDWT technique and enhancement SNR
Shahriar et al. Image reconstruction via compressed sensing
Nahar et al. A Novel Improve and Compression for the Medical Image Technique Based on the Double Density Wavelet

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20211012