An underwater noise-resistant video coding method, system and storage medium

Abstract

The application discloses an underwater noise-resistant video coding method and system and a storage medium. The method comprises an encoding method and a decoding method. The encoding method comprises the following steps of sequentially performing space-frequency domain conversion processing, graph region division processing, correlation detection processing, dynamic difference processing, source encoding processing and channel encoding processing on a digital video. The decoding method comprises the following steps of sequentially performing channel decoding processing, source decoding processing, prediction estimation processing, filtering estimation processing, video frame synthesis processing and frequency-space domain conversion processing on a received signal. The application can effectively compress a video data stream, reduce the influence of underwater noise on a video and realize real-time underwater high-quality video data transmission.

Description

Technical Field

[0001] This application relates to a video encoding and decoding method and system, belonging to the field of underwater communication technology, and particularly to an underwater noise-resistant video encoding and decoding method, system and storage medium. Background Technology

[0002] my country boasts a vast ocean area and abundant marine resources, including rich deposits of fish, minerals, oil, natural gas, and methane hydrate. In the exploration and exploitation of these resources, underwater vehicles or divers require underwater communication systems to transmit video signals. Furthermore, underwater communication technology is essential for marine environmental monitoring, deep-sea biological research, and national defense security.

[0003] There are three main underwater communication technologies: electromagnetic wave communication, acoustic wave communication, and optical wave communication. Electromagnetic wave communication suffers significant attenuation in water, making it suitable only for short-range communication; acoustic wave communication has a very low transmission rate; and optical wave communication can achieve high-speed information transmission. Therefore, underwater video transmission can only be achieved using optical wave communication.

[0004] The salinity, suspended particles, and planktonic flora and fauna in seawater have a significant impact on the underwater transmission of light signals, mainly through absorption, scattering, and turbulence. This generates considerable underwater noise, greatly reducing signal transmission quality. Current research has found that the wavelength range of 420-532nm, which has the lowest light loss and best light transmission in seawater, corresponds to the blue-green light band.

[0005] Currently, many countries are researching underwater optical communication technology, but it is still in the experimental stage and most have not yet been commercialized. Furthermore, video transmission systems for underwater blue-green optical communication are far from perfect and cannot guarantee high-speed transmission of real-time video signals. How to develop an effective video signal encoding and decoding technology for underwater blue-green optical communication networks, suppress underwater noise interference, and construct a real-time video transmission system is a current challenge and a research topic of significant application value. Summary of the Invention

[0006] According to one aspect of this application, an underwater noise-resistant video encoding and decoding method is provided, which can effectively compress video data streams and reduce the impact of underwater noise on video.

[0007] The underwater noise-resistant video encoding and decoding method described in this application includes an encoding method and a decoding method; wherein, the encoding method includes:

[0008] The video data is subjected to spatial-frequency domain transformation to obtain frequency domain data;

[0009] The frequency domain data is divided into image areas to obtain small-screen data;

[0010] The small screen data is subjected to correlation detection processing. If it is correlated, the small screen data does not need to be transmitted; otherwise, dynamic differential data is calculated.

[0011] After quantizing the dynamic differential data, source coding and channel coding are performed sequentially.

[0012] The decoding method includes:

[0013] The received signal is sequentially decoded into channel and then into source.

[0014] Calculate the dynamic differential data estimate based on the decoded data from the source;

[0015] The estimated values ​​of the dynamic differential data are filtered to calculate the estimated values ​​of the small-screen data;

[0016] Calculate the frequency domain data estimate based on the small-screen data estimate;

[0017] The frequency domain data estimate is converted to a frequency-space domain, and the video data estimate is calculated.

[0018] The estimated value of the video data is sent to the video display.

[0019] Furthermore, the spatial frequency domain conversion process includes:

[0020] Obtain video data S t (x, y), where the x-coordinate of a pixel in the video data is x = 0, 1, 2, ..., n-1, where n is the number of rows in a frame; the y-coordinate is y = 0, 1, 2, ..., m-1, where m is the number of columns in a frame; the sampling sequence number of the video signal is t = 0, 1, 2, ..., T, where T represents the duration of a segment of the video signal;

[0021] For video data S t The spatial domain to frequency domain transformation of (x, y) is performed using the discrete cosine transform method, yielding the following frequency domain data:

[0022]

[0023]

[0024]

[0025] Furthermore, the map area division process includes:

[0026] The frequency domain data D generated after spatial frequency domain transformation t If (u, v) is divided into N small frame data of a given type, then the i-th small frame data is represented as:

[0027] d t，i (u, v), i=1, 2, 3,...N, u=0, 1, 2,..., n1-1, v=0, 1, 2,..., m1-1, n1≤n, m1≤m;

[0028] Define the type partitioning threshold as:

[0029]

[0030] Where 0 < r ≤ 1; t is the sampling sequence number of the video signal, t = 0, 1, 2, ..., T, and T represents the duration of a segment of video signal.

[0031] Preferably, the given type includes three image types: high dynamic range image, low dynamic range image, and static image.

[0032] Preferably, the r values ​​are selected as λ1 and λ2, which respectively obtain the thresholds for judging the three types of images.

[0033] When the following conditions are met:

[0034]

[0035] The data type of this small image is a high-dynamic image;

[0036] When the following conditions are met:

[0037]

[0038] The data type of this small image is a weak dynamic image;

[0039] When the following conditions are met:

[0040]

[0041] The data type of this small image is static image.

[0042] Furthermore, the relevant detection processing includes:

[0043] The small screen data d obtained after the map area division process t，i (u, v) uses the average hash algorithm for relevant detection processing, and d t，i (u, v) and d t-1，i Perform relevant calculations on (u, v) to determine whether they are correlated.

[0044] Preferably, the specific calculation method for the related detection processing includes:

[0045] (1) Let the data of the i-th small frame be d. t，i (u, v) = [RGB] TWhere i = 1, 2, 3, ... N, t is the sampling sequence number of the video signal, t = 0, 1, 2, ..., T, T represents the duration of a video signal segment, and R, G, B represent the three primary color values ​​of the pixel;

[0046] definition:

[0047]

[0048] Put the data Convert to M-level grayscale value

[0049] (2) Calculate the average value of all pixels:

[0050]

[0051] Where u = 0, 1, 2, ..., n1-1, v = 0, 1, 2, ..., m1-1;

[0052] (3) Calculate the fingerprint of the small screen:

[0053] h t，i =[h t，i (0,0) h t，i (0, 1)...h t，i (n1-1, m1-1)] T ,

[0054] in,

[0055]

[0056] (4) Calculate the Hamming distance between two consecutive images:

[0057]

[0058] (5) If condition k is satisfied: t，i If the value is less than λ3, the two consecutive images are considered correlated; otherwise, they are considered uncorrelated. The value of parameter λ3 can be determined based on the video transmission quality requirements. The higher the video transmission quality requirements, the smaller the value of λ3, which is generally 5%-15% of the product of n1 and m1.

[0059] Furthermore, the calculation of the dynamic difference data is shown in the following formula:

[0060]

[0061]

[0062] Where, d t，i (u, v) represents the small-screen data at time t. This represents the dynamic difference data at time t, where u and v represent the horizontal and vertical coordinates of the data, respectively.

[0063] Preferably, the quantization of the dynamic difference data is: quantizing the dynamic difference data... Perform quantization with L-bit resolution. Preferably, L = 8.

[0064] Preferably, the source coding employs the Huffman coding method.

[0065] Preferably, the channel coding adopts the (g, k) Hamming code coding method.

[0066] Further, the calculation of frequency domain data estimates based on the small-screen data estimates includes:

[0067] Calculate small screen data d t，i Estimates of (u, v) The following formula:

[0068]

[0069] in, d represents the data of the i-th small frame at time t. t，i The estimated values ​​of (u, v) are i = 1, 2, 3, ... N, where u and v represent d. t，i The x and y coordinates of (u, v) This represents the estimated value of the dynamic difference data;

[0070] The frequency domain data D is calculated through video frame synthesis processing. t Estimates of (u, v) As shown in the formula below:

[0071]

[0072] Furthermore, the frequency-spatial domain transformation includes:

[0073] Through frequency-space domain transformation processing, based on frequency domain data D t Estimates of (u, v) Calculate the video data S t Estimate of (x, y) The following formula:

[0074]

[0075]

[0076]

[0077] Where u and v represent the frequency domain data D, respectively. tThe x and y coordinates of (u, v); t = 0, 1, 2, ..., T represents the sampling sequence number of the video signal, and T represents the duration of a segment of the video signal.

[0078] According to another aspect of this application, an underwater noise-resistant video encoding and decoding system is provided, the system comprising: a visual detector, a video encoder, an optical communication transmitter, an underwater channel, an optical communication receiver, a video decoder, and a video display connected in sequence; the video encoder is used to implement the algorithm of the above encoding method; the video decoder is used to implement the algorithm of the above decoding method; the optical communication transmitter and the optical communication receiver are based on blue-green laser communication.

[0079] According to another aspect of this application, a computer storage medium is provided that stores a computer program thereon, which, when executed by a processor, implements the steps of the above-described underwater noise-resistant video encoding and decoding method.

[0080] The beneficial effects that this application can produce include:

[0081] This application provides a video data encoding and decoding method and transmission technology for underwater blue-green optical communication systems. The method includes an encoding method and a decoding method. The encoding method includes sequentially performing spatial-frequency domain conversion processing, image area division processing, correlation detection processing, dynamic differential processing, source coding processing, and channel coding processing on digital video. The decoding method includes sequentially performing channel decoding processing, source decoding processing, prediction estimation processing, filtering estimation processing, video frame synthesis processing, and frequency-spatial domain conversion processing on the received signal. Because images have a high degree of similarity between adjacent pixels in the spatial domain, this application differs from existing technologies by using a spatial-frequency domain conversion process to remove this similarity and performing quantization encoding on dynamic differential data, thereby effectively reducing redundancy and improving encoding efficiency. In addition, the classification of images into three types (strong dynamic images, weak dynamic images, and static images) and the adoption of different encoding strategies for different types of images further improves encoding efficiency and video signal transmission quality, effectively reduces the impact of underwater noise on video image encoding and decoding, improves the compression ratio, and enables real-time underwater video transmission. This solves the technical challenges of underwater video transmission and is particularly suitable for applications such as underwater resource exploration, marine environmental monitoring, and submersible communication. Attached Figure Description

[0082] Figure 1 This is a flowchart of the underwater noise-resistant video encoding and decoding method described in this invention;

[0083] Figure 2 This is a schematic diagram of the underwater noise-resistant video encoding and decoding system described in this invention.

[0084] Figure 3This is a block diagram of a video encoder structure provided in an embodiment of the present invention;

[0085] Figure 4 This is a block diagram of a video decoder structure provided in an embodiment of the present invention. Detailed Implementation

[0086] The present application is described in detail below with reference to the embodiments, but the present application is not limited to these embodiments.

[0087] Please see Figure 1 The diagram illustrates a flowchart of the underwater noise-resistant video encoding and decoding method described in this application. The method includes an encoding method and a decoding method. The encoding method includes sequentially performing spatial-frequency domain conversion processing, image area division processing, correlation detection processing, dynamic differential processing, source coding processing, and channel coding processing on the digital video. The decoding method includes sequentially performing channel decoding processing, source decoding processing, prediction estimation processing, filtering estimation processing, video frame synthesis processing, and frequency-spatial domain conversion processing on the received signal.

[0088] The spatial frequency domain conversion process includes:

[0089] Step 1: Obtain video data S t (x, y), where (x, y) represents a pixel in the video data.

[0090] The x-coordinates are 0, 1, 2, ..., n-1;

[0091] The ordinate is y = 0, 1, 2, ..., m-1;

[0092] The sampling sequence numbers of the video signal are t = 0, 1, 2, ..., T.

[0093] Where n is the number of rows in a frame, m is the number of columns in a frame, and T represents the duration of a video signal.

[0094] Step 2: Process video data S t The spatial domain (x, y) is transformed to the frequency domain using the Discrete Cosine Transform (DCT) method, yielding the following frequency domain data:

[0095]

[0096]

[0097]

[0098] The map area division process includes:

[0099] The frequency domain data D generated after spatial frequency domain transformation t(u, v) is divided into N small frame data of three types, and the i-th data is represented as d. t，i (u, v), i=1, 2, 3,...N, u=0, 1, 2,..., 7, v=0, 1, 2,..., 7.

[0100] In one embodiment, n1 = 7, m1 = 7.

[0101] In one embodiment, the three types are: high dynamic range, low dynamic range, and static image; the type classification threshold is defined as:

[0102]

[0103] Where: 0 < r ≤ 1. The values ​​of r are selected as λ1 and λ2, for example: λ1 = 0.8, λ2 = 0.2, to obtain the thresholds for judging the three image types. Because image changes are affected by environmental noise and interference, λ1 and λ2 can be appropriately selected according to the specific environment. Generally, the value of λ1 can be set to 0.7–0.9, and the value of λ2 can be set to 0.1–0.4; when the condition is met:

[0104]

[0105] This small image data type is a high-dynamic image; when the following conditions are met:

[0106]

[0107] This small image data type is a low-motion image; when the following conditions are met:

[0108]

[0109] The data type of this small image is static image.

[0110] The relevant detection and processing include:

[0111] The small screen data d obtained after the map area division process t，i (u, v) uses the average hash algorithm for relevant detection processing, and d t，i (u, v) and d t-1，i Perform correlation calculations on (u, v) to determine if they are correlated. The specific calculation method is as follows:

[0112] (1) Let d t，i (u, v) = [RGB] T Where R, G, and B represent the three primary color values ​​of a pixel; Definition:

[0113]

[0114] Put the data Convert to M-level grayscale value In one embodiment, M is 256;

[0115] (2) Calculate the average value of all pixels:

[0116]

[0117] (3) Calculate the fingerprint of the small screen:

[0118] h t，i =[h t，i (0,0) h t，i (0, 1)...h t，i (7, 7)] T ,

[0119] in,

[0120]

[0121] (4) Calculate the Hamming distance between two consecutive images:

[0122]

[0123] (5) If condition k is satisfied: t，i If the score is less than 10, the two consecutive images are considered related; otherwise, they are considered unrelated.

[0124] The dynamic differential processing includes:

[0125] For data that has been confirmed as irrelevant through relevant testing and processing, dynamic difference data is calculated using the dynamic difference module:

[0126]

[0127] Where, d t-1，i (u, v) represents the small-frame data at time t-1, d t，i (u, v) represents the small-screen data at time t. This represents the dynamic difference data at time t.

[0128] The source coding process includes:

[0129] For dynamic difference data The signal is quantized to 8-bit resolution and then encoded using a source coding module. The encoding method used is Huffman coding.

[0130] The channel coding process includes: using (g, k) Hamming code encoding to encode the data after source coding, for example, g = 7, k = 4.

[0131] The decoding method includes:

[0132] Step 1: Decode the received signal through channel decoding processing;

[0133] Step 2: After channel decoding, the signal is then decoded through source decoding.

[0134] Step 3: After decoding the source data, calculate the dynamic difference data estimate through prediction estimation.

[0135] Step 4: Estimate the dynamic difference data through filtering estimation. Filter the image and calculate the small-screen data d. t，i The estimated value of (u, v) is given by the following formula:

[0136]

[0137] in, Represents the small-screen data d at time t. t，i The estimated value of (u, v) This represents the small-screen data d at time t-1. t，i Estimates of (u, v);

[0138] Step 5: Through video frame synthesis processing, based on small-screen data d t，i Estimates of (u, v) Calculate the frequency domain data D t The estimated value of (u, v) is given by the following formula:

[0139]

[0140] Step 6: Through frequency-space domain transformation processing, based on frequency domain data D t Estimates of (u, v) Calculate the video data S t Estimate of (x, y) The following formula:

[0141]

[0142]

[0143]

[0144] Step 7: Transfer video data S t The estimated value of (x, y) is sent to the video display for display.

[0145] The video transmission system based on an underwater noise-resistant video encoding and decoding method described in this invention, such as... Figure 2As shown, the system includes a visual detector, a video encoder, an optical communication transmitter, an underwater channel, an optical communication receiver, a video decoder, and a video display. The visual detector is connected to the video encoder, the video encoder is connected to the optical communication transmitter, the optical communication transmitter is connected to the underwater channel, the underwater channel is connected to the optical communication receiver, the optical communication receiver is connected to the video decoder, and the video decoder is connected to the video display.

[0146] In one implementation, the video encoder is implemented as follows: Figure 3 As shown, it includes:

[0147] Step 1: The visual detector uses a CCD imaging camera and its detection system to obtain video data S. t (x, y) is then passed to the spatial frequency domain conversion module in the video encoder, where

[0148] x = 0, 1, 2, ..., n-1

[0149] y = 0, 1, 2, ..., m-1

[0150] t = 0, 1, 2, ..., T

[0151] Where T represents the duration of a video signal, t represents the sampling sequence number of the video signal, and n and m represent the number of rows and columns of a frame, respectively;

[0152] Step 2: The spatial frequency domain conversion module converts the video data S t The spatial domain (x, y) is transformed to the frequency domain using the Discrete Cosine Transform (DCT) method, yielding the following frequency domain data:

[0153]

[0154]

[0155]

[0156] Step 3: Generate frequency domain data D through the spatial frequency domain conversion module. t After (u, v), the image area division module further divides it into N small image data of three types, with the i-th data represented as: d t，i (u, v), i = 1, 2, 3, ..., N, u = 0, 1, 2, ..., n1-1, v = 0, 1, 2, ..., m1-1. The three types include: strong motion scenes, weak motion scenes, and static scenes. The type classification threshold is defined as:

[0157]

[0158] Where: 0 < r ≤ 1. The values ​​of r are selected as λ1 and λ2, for example: λ1 = 0.8 and λ2 = 0.2, to obtain the thresholds for judging the three image types. When the condition is met:

[0159]

[0160] This small image data type is a high-dynamic image; when the following conditions are met:

[0161]

[0162] This small image data type is a low-motion image; when the following conditions are met:

[0163]

[0164] The data type of this small image is static image;

[0165] Step 4: Divide the small screen data d using the image area division module. t，i After (u, v), the relevant detection module uses the average hash algorithm to process d. t，i (u, v) and d t-1，i Perform relevant calculations on (u, v) and determine whether they are correlated. The specific calculation method is as follows:

[0166] (1) The small screen data d is divided by the image area division module. t，i After (u, v), the relevant detection module uses the average hash algorithm to process d. t，i (u, v) and d t-1，i Perform relevant calculations on (u, v) and determine whether they are correlated. The specific calculation method is as follows:

[0167] (1) Let d t，i (u, v) = [RGB] T Where R, G, and B represent the three primary color values ​​of a pixel; Definition:

[0168]

[0169] Put the data Convert to M-level grayscale value

[0170] (2) Calculate the average value of all pixels:

[0171]

[0172] (3) Calculate the fingerprint of the small screen:

[0173] h t，i =[h t，i (0,0) h t，i (0, 1)...ht，i (7, 7)] T ,

[0174] in,

[0175]

[0176] (4) Calculate the Hamming distance between two consecutive images:

[0177]

[0178] (5) If condition k is satisfied: t，i If the value is less than λ3, the two consecutive images are considered related; otherwise, they are considered unrelated. The value of parameter λ3 can be determined based on the video transmission quality requirements. The higher the video transmission quality requirements, the smaller the value of λ3, which is generally 5%-15% of the product of n1 and m1.

[0179] Step 5: After calculation by the relevant detection module, if the data is relevant, the small screen data does not need to be transmitted; if it is not relevant, the dynamic difference data is calculated by the dynamic difference module.

[0180]

[0181] Step 6: Process the dynamic difference data The signal is quantized to 8-bit resolution and then encoded by the source coding module using the Huffman coding method.

[0182] Step 7: After the source coding module completes the source coding, the channel coding module performs channel coding, using (7,4) Hamming code.

[0183] In one implementation, the video decoder is implemented as follows: Figure 4 As shown, it includes:

[0184] Step 1: The transmitted signal is received by the optical communication receiver and then decoded by the channel decoding module;

[0185] Step 2: After decoding by the channel decoding module, it is then decoded by the source decoding module;

[0186] Step 3: After decoding by the source decoding module, the dynamic difference data estimate is calculated by the prediction estimation module.

[0187] Step 4: Estimate the dynamic difference data using the filtering estimation module. Filter the image and calculate the small-screen data d. t，i The estimated value of (u, v) is given by the following formula:

[0188]

[0189] in, Represents the small-screen data d at time t. t，i The estimated value of (u, v) This represents the small-screen data d at time t-1. t，i Estimates of (u, v);

[0190] Step 5: Through video frame synthesis processing, based on small-screen data d t，i Estimates of (u, v) Calculate the frequency domain data D t The estimated value of (u, v) is given by the following formula:

[0191]

[0192] Step 6: Through frequency-space domain transformation processing, based on frequency domain data D t Estimates of (u, v) Calculate the video data S t Estimate of (x, y) The following formula:

[0193]

[0194]

[0195]

[0196] Step 7: Finally, transfer the video data S t The estimated value of (x, y) is sent to the video display for display.

[0197] This embodiment provides an underwater noise-resistant video encoding and decoding method and its transmission system. Utilizing an underwater blue-green optical communication system, it presents a video data encoding, decoding, and transmission technology that effectively reduces the impact of underwater noise and solves the technical challenges of underwater video transmission.

[0198] It should be pointed out that there are many parameters that can be selected and adjusted in the encoding and decoding method, such as: T, N, L, M, n, m, n1, m1, λ1, λ2, λ3. The adjustment or change of these parameters does not change the idea of ​​this invention and should also be within the scope of protection of this invention.

[0199] The above description is merely a few embodiments of this application and is not intended to limit this application in any way. Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any changes or modifications made by those skilled in the art without departing from the scope of the technical solution of this application using the disclosed technical content are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for encoding and decoding underwater noise-resistant video, characterized in that, The method includes an encoding method and a decoding method; wherein, the encoding method includes: The video data is subjected to spatial-frequency domain transformation to obtain frequency domain data; The spatial frequency domain conversion process includes: Obtain video data S t ( x,y ), where the x-coordinate of a pixel in the video data. x =0,1,2,..., n -1, n The vertical axis represents the number of rows in a single frame of an image; the vertical axis represents the number of rows in a single frame of an image y =0,1,2,..., m -1, m The column number of a single frame; the sequence number of the video signal samples. t =0,1,2,..., T,T Indicates the duration of a video signal; For video data S t ( x,y Perform a spatial domain to frequency domain conversion to obtain the frequency domain data: in, u and v Representing frequency domain data respectively D t (u,v) The x and y coordinates; The frequency domain data is divided into image areas to obtain small-screen data; The small screen data is subjected to correlation detection processing. If it is correlated, the small screen data does not need to be transmitted; otherwise, dynamic differential data is calculated. After quantizing the dynamic differential data, source coding and channel coding are performed sequentially. The decoding method includes: The received signal is sequentially decoded into channel and then into source. Calculate the dynamic differential data estimate based on the decoded data from the source; The estimated values ​​of the dynamic differential data are filtered to calculate the estimated values ​​of the small-screen data; Calculate the frequency domain data estimate based on the small-screen data estimate; The frequency domain data estimate is converted to a frequency-space domain, and the video data estimate is calculated. The estimated value of the video data is sent to the video display.

2. The underwater noise-resistant video encoding and decoding method according to claim 1, characterized in that, The map area division process includes: Frequency domain data generated after spatial frequency domain transformation D t ( u,v ) divided into N Given a small image data of a certain type, then the first... i The data for each small screen is represented as follows: d t,i ( u,v ), i =1,2,3,… N , u =0,1,2,...,n1-1,, v =0,1,2,..., m 1-1, n 1≤ n , m 1≤ m ; Define the type partitioning threshold as: in, u and v Representing frequency domain data respectively D t ( u,v The x and y coordinates of ) r This represents the coefficient selected when defining the type partitioning threshold, satisfying the condition: 0 <r≤1; t This represents the sequence number of the video signal sampling. t =0,1,2,...,T,T represents the duration of a video signal.

3. The underwater noise-resistant video encoding and decoding method according to claim 2, characterized in that, The given type includes three image types: high dynamic range, low dynamic range, and static image.

4. The underwater noise-resistant video encoding and decoding method according to claim 3, characterized in that, r The threshold values ​​λ1 and λ2 are selected to determine the three types of images, where the value of λ1 is set to 0.7 to 0.9 and the value of λ2 is set to 0.1 to 0.

4. When the following conditions are met: The data type of this small image is a high-dynamic image; When the following conditions are met: The data type of this small image is a weak dynamic image; When the following conditions are met: The data type of this small image is static image.

5. The underwater noise-resistant video encoding and decoding method according to claim 1, characterized in that, The relevant detection and processing include: The small screen data divided by the map area division process d t,i ( u,v The average hash algorithm is used for relevant detection and processing to determine... d t,i ( u,v )and d t-1,i ( u,v Is it relevant? u and v Representing frequency domain data respectively D t ( u,v The x and y coordinates of (). t =0,1,2,...,T represents the sampling sequence number of the video signal. T Indicates the duration of a video signal; i =1,2,3,… N This refers to the sequence number of the small screen data.

6. The underwater noise-resistant video encoding and decoding method according to claim 5, characterized in that, The specific calculation method for the relevant detection processing includes: (1) Set small screen data d t,i ( u,v ) = [RGB] T Where R, G, and B represent small-screen data. d t,i ( u,v The values ​​of the three primary colors; definition: Put the data Convert to M-level grayscale value ; (2) Calculate the average value of all pixels: in, u =0,1,2,...,n1-1, v =0,1,2,..., m 1-1; (3) Calculate the fingerprint of the small screen: ， in, (4) Calculate the Hamming distance between two consecutive images: (5) If the condition is met: K t,i < λ 3. If two consecutive images are related, they are considered related; otherwise, they are considered unrelated. λ 3 is a given parameter.

7. The underwater noise-resistant video encoding and decoding method according to claim 6, characterized in that, λ 3 is 5% to 15% of the product of n1 and m1.

8. The underwater noise-resistant video encoding and decoding method according to claim 1, characterized in that, The calculation of the dynamic difference data is shown in the following formula: in, d t,i ( u,v () represents the small-screen data at time t. This represents the dynamic difference data at time t. u and v These represent the horizontal and vertical coordinates of the data, respectively. The quantization of the dynamic difference data is as follows: For the dynamic difference... Quantify the data; The source coding employs the Huffman coding method; The channel coding adopts ( g,k Hamming code encoding method.

9. The underwater noise-resistant video encoding and decoding method according to claim 1, characterized in that, The calculation of the frequency domain data estimate based on the small-screen data estimate includes: Calculate small screen data d t,i ( u,v The estimated value of ) The formula is as follows: in, express t Time of the first i Small screen data d t,i ( u,v The estimated value of ) i =1,2,3,… N , u and v They represent d t,i ( u,v The x and y coordinates of ) This represents the estimated value of the dynamic difference data; Frequency domain data is calculated through video frame synthesis processing. D t ( u,v The estimated value of ) As shown in the formula below: 。 10. The underwater noise-resistant video encoding and decoding method according to claim 1, characterized in that, The frequency-spatial domain conversion includes: Through frequency-space domain transformation processing, based on frequency domain data D t ( u,v The estimated value of ) Calculate the video data S t ( x,y The estimated value of ) The formula is as follows: in, u and v Representing frequency domain data respectively D t ( u,v The x and y coordinates of (). t =0,1,2,...,T, represents the sequence number of the video signal sampling. T Indicates the duration of a video signal; m is the number of columns in a frame of an image; n is the number of rows in a frame of an image.

11. A video encoding and decoding system resistant to underwater noise, characterized in that, The system includes: a visual detector, a video encoder, an optical communication transmitter, an underwater channel, an optical communication receiver, a video decoder, and a video display connected in sequence; the video encoder is used to implement the encoding step of the encoding and decoding method as described in any one of claims 1-8; the video decoder is used to implement the decoding step of the encoding and decoding method as described in any one of claims 1, 9-10; the optical communication transmitter and the optical communication receiver are based on blue-green laser communication.

12. A computer storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the underwater noise-resistant video encoding / decoding method according to any one of claims 1-10.

Images