Satellite remote sensing-based salt tide real-time monitoring method and system

Abstract

The invention belongs to the technical field of image communication, and particularly relates to a salt tide real-time monitoring method and system based on satellite remote sensing, and the method comprises the steps: collecting a remote sensing image of a monitoring region in real time; allocating code words for pixel values in each channel of the remote sensing image by using a variable-length coding algorithm; dividing the code word of each pixel value into a plurality of layers according to the loss threshold value of each layer to obtain a sub-code word of each layer of each pixel value; and performing layered coding on each channel of the remote sensing image by using the sub-code words of each layer of each pixel value, and transmitting coding results of different layers to a monitoring terminal in sequence for salt moisture real-time monitoring. Layered compression transmission is performed on the remote sensing image, and the real-time performance of salt tide monitoring is improved.

Description

Technical Field

[0001] The present invention relates to the field of image communication technology, and more specifically, to a method and system for real-time monitoring of salt tide based on satellite remote sensing. Background Art

[0002] Salt intrusion is a hydrological phenomenon in which seawater flows back into the estuary driven by tides, causing the salinity of freshwater to increase. Its essence is a special interface movement formed by the interaction between hydrological dynamics and salinity gradient. This phenomenon often occurs in estuary areas in winter (such as the Yangtze River Estuary and the Pearl River Estuary). Salt intrusion can cause abnormal increases in freshwater salinity, directly threatening the safety of drinking water for residents. It may also aggravate soil salinization and increase the cost of agricultural desalination and industrial equipment corrosion prevention. Therefore, it is necessary to monitor salt intrusion in real time so that timely countermeasures can be taken.

[0003] Real-time monitoring of salt tide intrusion relies on dynamic data of water salinity with high temporal and spatial resolution. Traditional monitoring methods mainly use ground stations and buoy sensors to infer salinity distribution through parameters such as conductivity and water level. However, this monitoring method has limited spatial coverage and high maintenance costs, making it difficult to meet the continuous observation needs of large estuary areas. Satellite remote sensing technology has become an important means of salt tide monitoring due to its wide-area coverage advantage.

[0004] However, the massive amount of remote sensing images collected by satellite remote sensing technology will bring great pressure on transmission and storage, so it is necessary to compress the remote sensing images. Although existing compression technologies (such as JPEG2000, deep learning autoencoders, etc.) can achieve high compression ratios, they may lose detailed information about salt tide intrusion.

[0005] Therefore, there is an urgent need for a compression technology for remote sensing images to retain the detailed information of salt intrusion in remote sensing images and at the same time ensure the real-time monitoring of salt intrusion. Summary of the invention

[0006] In order to solve the technical problem that detailed information of salt tide intrusion may be lost when compressing remote sensing images using existing compression methods, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a method for real-time monitoring of salt tide based on satellite remote sensing, comprising: The remote sensing images of the monitoring area are collected in real time; codewords are assigned to the pixel values ​​in each channel of the remote sensing images using a variable-length coding algorithm; the codewords of each pixel value are divided into several layers according to the loss threshold of each layer to obtain the sub-codewords of each layer of each pixel value; the sub-codewords of each layer of each pixel value are used to perform hierarchical coding on each channel of the remote sensing image, and the coding results of different layers are transmitted to the monitoring terminal in sequence for real-time monitoring of salt tide.

[0008] The present invention intelligently hierarchizes the codewords of each pixel value according to a preset loss threshold, generates a multi-level sub-codeword structure, performs hierarchical encoding processing on each channel, and sequentially transmits the encoded data to the monitoring terminal according to the hierarchical priority through the network. The hierarchical encoding mechanism ensures the priority transmission of key data, maintains the real-time performance of the monitoring system under limited bandwidth conditions, and at the same time the scalable hierarchical structure allows dynamic adjustment of the transmission strategy according to the network conditions, ensuring the integrity and timeliness of the salt tide monitoring data and providing reliable data support for disaster prevention decision-making.

[0009] Preferably, the method for obtaining the sub-codewords of each layer of each pixel value is as follows: S101: Use the pixel values whose first bit positions of the codeword are the same as the first bit positions of the codeword of the target pixel value as the homomorphic pixel values of the target pixel value at the th layer. is the total length of the sub-codewords of the first layers of the target pixel value, is the segmentation length of the th layer; S102: Determine the th loss degree when the target bit is used as the sub-codeword of the th layer of the target pixel value and its homomorphic pixel values at the th layer. The target bit is the th to the th bit positions of the codeword of the target pixel value; S103: In response to the th loss degree not being greater than the preset loss threshold of the th layer, use the target bit as the sub-codeword of the th layer of the target pixel value and its homomorphic pixel values at the th layer. Otherwise, increment the value of by one, and repeat steps S101 - S103 until the sub-codeword of the th layer of the target pixel value and its homomorphic pixel values at the th layer is obtained and the iteration stops.

[0010] By introducing a loss threshold independent between layers, the present invention improves the traditional fixed coding mode into a dynamic adjustment process, enabling the sub-codewords of each layer of pixel values to meet the loss threshold requirements, thereby ensuring that the quality of the image restored according to the encoding results of each layer during decoding can meet the salt tide monitoring requirements.

[0011] Preferably, the th loss degree satisfies the expression: ; where represents the th loss degree when the target bit is used as the sub-codeword of the th layer of the target pixel value and its homomorphic pixel values at the Degree of loss represents the number of homomorphic pixel values of the layer corresponding to the target pixel value; represents the th homomorphic pixel value of the layer corresponding to the target pixel value; represents the target pixel value; represents the number of pixel points corresponding to the th homomorphic pixel value of the layer corresponding to the target pixel value.

[0012] In the present invention, by calculating the degree of loss, the system quantifies the information loss under different coding granularities, providing a decision basis for the sub-codeword allocation of each layer. In the process of obtaining the degree of loss, the present invention considers the proportion of the number of pixel points corresponding to the homomorphic pixel values, effectively eliminating the statistical deviation caused by uneven pixel distribution and improving the robustness of the loss degree measurement.

[0013] Preferably, the hierarchical coding of each channel of the remote sensing image by using the sub-codewords of each layer of each pixel value includes: for each pixel point in each channel of the remote sensing image, using the sub-codewords of each layer corresponding to the pixel value of the pixel point as the coding result of the pixel point in each layer; splicing the coding results of all pixel points in the same layer together in sequence as the coding result of each channel of the remote sensing image in this layer.

[0014] In the present invention, by hierarchically coding each channel of the remote sensing image, the priority transmission of key data in the remote sensing image is ensured, and the real-time performance of the salt tide system can still be maintained under limited bandwidth conditions.

[0015] Preferably, the allocation of codewords to the pixel values in each channel of the remote sensing image by using the variable-length coding algorithm includes: obtaining the frequencies of the pixel values in the remote sensing image, and using Huffman coding to obtain the codewords of the pixel values according to the frequencies of the pixel values.

[0016] Preferably, the allocation of codewords to the pixel values in each channel of the remote sensing image by using the variable-length coding algorithm includes: obtaining the frequencies of the pixel values in the remote sensing image, sorting the pixel values according to the frequencies of the pixel values, and using Shannon-Fano coding to obtain the codewords of the pixel values according to the sorted order of the pixel values and the frequencies of the pixel values.

[0017] Preferably, the sorting of the pixel values according to the frequencies of the pixel values includes: sorting the pixel values in descending order according to the frequencies of the pixel values.

[0018] Preferably, sorting the pixel values according to the frequency of each pixel value includes: S201: constructing an empty sequence, denoted as the pixel sequence; S202: adding the pixel value with the highest frequency to the pixel sequence; S203: for all pixel values not added to the pixel sequence, determining the priority of the pixel value according to the frequency of the pixel value and the difference between the pixel value and the last pixel value in the current pixel sequence: , represents the priority of the th pixel value not added to the pixel sequence currently; represents the th pixel value not added to the pixel sequence currently; represents the last pixel value in the current pixel sequence; represents the th pixel value not added to the pixel sequence currently; represents the absolute value symbol; adding the pixel value with the highest priority to the end of the pixel sequence; S204: repeating S203 until all pixel values are added to the pixel sequence and then stopping the iteration.

[0019] Based on the frequency of pixel values, the present invention determines the priority of pixel values to be added to the pixel sequence, so that the pixel sequence effectively retains the main statistical features of each channel of the remote sensing image, ensuring that the codewords of high-frequency pixel values are as short as possible, thereby improving the coding efficiency. The present invention takes into account the differences between pixel values on the basis of frequency, thereby calculating the priority of pixel values to be added to the pixel sequence, so that similar pixel values in the pixel sequence are arranged together as much as possible. Then, in the codewords of each pixel value obtained by Shannon-Fano coding, the codewords of similar pixel values are also similar. When layering the codewords, similar pixel values in the lower layer correspond to the same shorter sub-codewords, thereby reducing the amount of compressed data in the lower layer, and thus improving the real-time performance of salt tide monitoring.

[0020] Preferably, it further includes: in response to the monitoring terminal receiving the coding results of the first layer of each channel of the remote sensing image, decompressing the coding results of the first layer of each channel to obtain a decompressed image, and identifying whether there is a salt tide intrusion according to the decompressed image; in response to the monitoring terminal receiving the coding results of the remaining layers of each channel of the remote sensing image and there being a salt tide intrusion, decompressing the coding results of the remaining layers in sequence, updating the decompressed image according to the decompression results, and obtaining the detailed information of the salt tide intrusion according to the finally obtained decompressed image.

[0021] Through hierarchical processing, the present invention improves the response speed of salt tide monitoring. The monitoring system preferentially decompresses and analyzes the first-layer encoded data to quickly complete the preliminary judgment of salt tide intrusion. This fast screening mechanism based on the key layer significantly reduces the consumption of computing resources, enabling the salt tide monitoring system to quickly complete the preliminary warning. Only when it is confirmed that there is a salt tide intrusion, the system continues to decompress the data of the remaining levels. This condition-triggered processing strategy avoids unnecessary data decompression operations and effectively saves computing resources.

[0022] In a second aspect, the present invention provides a real-time salt tide monitoring system based on satellite remote sensing, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned real-time salt tide monitoring method based on satellite remote sensing is implemented.

[0023] By adopting the above technical solution, the above-mentioned real-time salt tide monitoring method based on satellite remote sensing is generated into a computer program and stored in the memory to be loaded and executed by the processor. Thus, a terminal device is manufactured according to the memory and the processor, which is convenient to use.

[0024] The beneficial effects of the present invention are as follows: The hierarchical coding mechanism of the present invention ensures the priority transmission of key data, maintains the real-time performance of the monitoring system under limited bandwidth conditions, improves the response speed of salt tide monitoring, and at the same time, the scalable hierarchical structure allows dynamic adjustment of the transmission strategy according to the network conditions, ensuring the integrity and timeliness of salt tide monitoring data and providing reliable data support for disaster prevention decision-making; the present invention preferentially decompresses and analyzes the first-layer encoded data to quickly complete the preliminary judgment of salt tide intrusion, and only when it is confirmed that there is a salt tide intrusion, it continues to decompress the data of the remaining levels to obtain the details of salt tide intrusion, avoiding unnecessary data decompression operations and effectively saving computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flowchart schematically showing a real-time salt tide monitoring method based on satellite remote sensing in the present invention; Figure 2 is a schematic diagram showing a Huffman tree; Figure 3 is schematically showing Figure 2 the codewords of each pixel value corresponding to the Huffman tree in; Figure 4 is a schematic diagram showing a Shannon-Fano tree constructed in the order of decreasing pixel value frequency; Figure 5 is schematically showing Figure 4 the codewords of each pixel value corresponding to the Shannon-Fano tree in; Figure 6is a schematic diagram showing a Shannon-Fano tree constructed according to the order of pixel values in a pixel sequence; Figure 7 is a schematic illustration showing Figure 6 the codewords of the respective pixel values corresponding to the Shannon-Fano tree in; Figure 8 is a schematic illustration showing for Figure 3 the sub-codewords of each layer of the respective pixel values obtained by hierarchically dividing the codewords of the respective pixel values in; Figure 9 is a schematic illustration showing for Figure 5 the sub-codewords of each layer of the respective pixel values obtained by hierarchically dividing the codewords of the respective pixel values in; Figure 10 is a schematic illustration showing for Figure 7 the sub-codewords of each layer of the respective pixel values obtained by hierarchically dividing the codewords of the respective pixel values in; Figure 11 is a schematic illustration showing according to Figure 8 the decoding results of each layer obtained by decoding the encoding results of each layer with the sub-codewords of each layer of the respective pixel values in; Figure 12 is a schematic illustration showing according to Figure 9 the decoding results of each layer obtained by decoding the encoding results of each layer with the sub-codewords of each layer of the respective pixel values in; Figure 13 is a schematic illustration showing according to Figure 10 the decoding results of each layer obtained by decoding the encoding results of each layer with the sub-codewords of each layer of the respective pixel values in. Detailed implementation manners

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Next, the detailed implementation manners of the present invention will be described in detail in conjunction with the accompanying drawings.

[0028] An embodiment of the present invention discloses a real-time salt tide monitoring method based on satellite remote sensing. Referring to Figure 1 , it includes steps S001 - step S006: S001. Real-time collect remote sensing images of the monitoring area.

[0029] Specifically, satellite remote sensing technology is used to collect remote sensing images of the monitoring area in real time. In this embodiment, the collected remote sensing images are visible light images, that is, RGB images. In other embodiments, the collected remote sensing images can be infrared images, hyperspectral images, etc.

[0030] S002. Use the variable-length coding algorithm to assign codewords to the pixel values in each channel of the remote sensing image.

[0031] The remote sensing image is an RGB image, including three channels of R, G, and B. In the present invention, the images corresponding to each channel of the remote sensing image are encoded separately. Therefore, in the present invention, the image of any one of the R channel, G channel, and B channel is used as the target image, and the specific method for encoding the images corresponding to each channel of the remote sensing image is described by taking the target image as an example.

[0032] In the first embodiment, the variable-length coding algorithm uses Huffman coding. Then, Huffman coding is used to assign codewords to the pixel values in the target image, including: Obtain the frequencies of the pixel values in the target image, and use Huffman coding to obtain the codewords of the pixel values according to the frequencies of the pixel values.

[0033] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, the pixel values include 19, 20, 21, 22, 23, 24, 25, 77, 78, 79, 80, 189, 190, 192, 193 respectively, and the frequencies corresponding to the pixel values are 、 、 、 、 、 、 、 、 、 、 、 、 、 、 , the Huffman tree constructed according to the frequencies of the pixel values is shown in Figure 2 , and the codewords of the pixel values are shown in Figure 3 .

[0034] In the second embodiment, the variable-length coding algorithm adopts Shannon–Fano coding. Then, the Shannon–Fano coding is used to assign codewords to the pixel values in the target image, including: Obtain the frequencies of the pixel values in the target image, sort the pixel values in descending order according to the frequencies of the pixel values, and use the Shannon–Fano coding to obtain the codewords of the pixel values according to the sorted order of the pixel values and the frequencies of the pixel values.

[0035] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, the pixel values include 19, 20, 21, 22, 23, 24, 25, 77, 78, 79, 80, 189, 190, 192, 193, and the corresponding frequencies of the pixel values are , , , , , , , , , , , , , , . The result of sorting the pixel values in descending order according to the frequencies of the pixel values is {80, 23, 190, 79, 22, 192, 78, 20, 189, 77, 19, 21, 24, 25, 193}. For a schematic diagram of the Shannon–Fano tree constructed in descending order of the pixel value frequencies, see Figure 4 , and the codewords of the pixel values can be seen in Figure 5 .

[0036] In the third embodiment, the variable-length coding algorithm adopts Shannon–Fano coding. Obtain the frequencies of the pixel values in the target image, sort the pixel values according to the frequencies of the pixel values and the differences between the pixel values, and use the Shannon–Fano coding to obtain the codewords of the pixel values according to the sorted order of the pixel values and the frequencies of the pixel values. Among them, sorting the pixel values according to the frequencies of the pixel values and the differences between the pixel values includes: Construct an empty sequence, denoted as the pixel sequence. Add the pixel value with the highest frequency to the pixel sequence. For all pixel values not yet added to the pixel sequence, determine the priority of each pixel value based on its frequency and the difference between this pixel value and the last pixel value in the current pixel sequence. Then add the pixel value with the highest priority to the end of the pixel sequence to achieve one update of the pixel sequence. After each update of the pixel sequence, re-determine the priorities of all pixel values not yet added to the pixel sequence, and add the pixel value with the highest priority to the end of the pixel sequence to achieve another update of the pixel sequence. Repeat the above process until all pixel values have been added to the pixel sequence, at which point the iteration stops. The finally obtained pixel sequence is the result of sorting the pixel values in the target image.

[0037] Among them, the priority of the pixel value not yet added to the pixel sequence satisfies the expression: ; Where represents the priority of the -th pixel value not yet added to the pixel sequence currently; represents the -th pixel value not yet added to the pixel sequence currently; represents the last pixel value in the current pixel sequence; represents the frequency of the -th pixel value not yet added to the pixel sequence currently; represents the absolute value symbol.

[0038] It should be noted that the purpose of the present invention is for the hierarchical compression and transmission of remote sensing images. The compressed data at the lower layer contains the global information in the remote sensing image, and the compressed data at the higher layer contains the detailed information in the remote sensing image. Therefore, in the third embodiment, sorting the pixel values according to the frequency of the pixel values and the differences between the pixel values, firstly, can ensure that the codewords of the pixel values with higher frequencies are as short as possible, thus guaranteeing the compression efficiency; secondly, the longer the same part of the codewords assigned to similar pixel values is, so that when performing hierarchical compression according to the codewords of the pixel values subsequently, at the lower layer, more similar pixel values can use the same part of the codewords for encoding, to achieve a small loss degree at the lower layer while the data volume corresponding to the compression result is small, thereby ensuring the real-time monitoring of salt tides.

[0039] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, the pixel values include 19, 20, 21, 22, 23, 24, 25, 77, 78, 79, 80, 189, 190, 192, 193, and the corresponding frequencies of each pixel value are , , , , , , , , , , , , , , . Add the pixel value 80 with the maximum frequency to the pixel sequence, then the pixel sequence is {80}, and the priority of the pixel value 19 is , and the priority of the pixel value 20 is . Similarly, calculate the priorities of the pixel values 21, 22, 23, 24, 25, 77, 78, 79, 189, 190, 192, 193 respectively. Among them, the priority of the pixel value 79 is the largest, which is . Therefore, add the pixel value 79 to the end of the pixel sequence, then the pixel sequence is {80, 79}; similarly, according to the new pixel sequence, calculate the priorities of the pixel values that have not been added to the pixel sequence again, and add the pixel value with the largest priority to the end of the pixel sequence, and repeat the above process continuously. Finally, the obtained pixel sequence is {80, 79, 78, 77, 23, 22, 21, 20, 19, 24, 25, 190, 189, 192, 193}. For the schematic diagram of the Shannon-Fano tree constructed according to the order of the pixel values in the pixel sequence, see Figure 6 , and for the codewords of the pixel values, see Figure 7 .

[0040] S003. Divide the codewords of each pixel value into several layers according to the loss thresholds of each layer to obtain the sub-codewords of each layer of each pixel value.

[0041] Specifically, starting from the pixel value with the highest frequency, obtain the first layer in the codewords of each pixel value as the sub-codewords of the first layer of each pixel value; in response to the sub-codewords of the first layer of the pixel value being the same as the codeword of the pixel value, there are no sub-codewords of the first layer for this pixel value, otherwise, obtain the second layer in the codewords of each pixel value based on the sub-codewords of the first layer as the sub-codewords of the second layer of each pixel value; in response to the bit string formed by concatenating the sub-codewords of the first layer and the second layer of the pixel value being the same as the codeword of the pixel value, there are no sub-codewords of the third layer for this pixel point, otherwise, obtain the third layer in the codewords of each pixel value based on the sub-codewords of the first layer and the sub-codewords of the second layer as the sub-codewords of the third layer of each pixel value; and so on, until there are no sub-codewords of the next layer for all pixel values, then stop the iteration.

[0042] Taking the th layer ( ), for example, illustrate the method of obtaining the sub-codewords of the th layer of each pixel value: 1. When , take the pixel value with the highest frequency among all pixel values that have not obtained the sub-codewords of the th layer as the target pixel value; when , in response to the bit string formed by concatenating the sub-codewords of the previous layers of the pixel value in the order of the layers being different from the codeword of the pixel value, take the pixel value as a candidate pixel value, and take the candidate pixel value with the highest frequency among all candidate pixel values that have not obtained the sub-codewords of the th layer as the target pixel value.

[0043] 2. Set the segmentation length of the th layer, and the initial value of it is 1.

[0044] 3. When , take the first bit positions of the codeword of the target pixel value as the target bits, and obtain the pixel values whose first bit positions are the same as the target bits of the target pixel value as the homomorphic pixel values of the th layer of the target pixel value; when , take the th bit position to the th bit position of the codeword of the target pixel value as the target bits, and obtain the pixel values whose first bit positions are the same as the first bit positions of the target pixel value as the homomorphic pixel values of the th layer of the target pixel value, where represents the length of the sub-codewords of the th layer of the target pixel value, Indicates the total length of the sub-codewords of the previous layer of the target pixel value.

[0045] 4. Determine the th layer of the target bit as the target pixel value and the difference between each homomorphic pixel value of its th layer, as well as the number of pixel points corresponding to each homomorphic pixel value, to determine the th layer of the sub-codeword when the target bit is used as the target pixel value and the th layer of the homomorphic pixel value of its ; Among them, indicates the th layer of the sub-codeword when the target bit is used as the target pixel value and the th layer of the homomorphic pixel value of its th layer, indicates the number of homomorphic pixel values of the th layer of the target pixel value; indicates the th layer of the th homomorphic pixel value of the target pixel value; indicates the target pixel value; indicates the th layer of the th homomorphic pixel value of the target pixel value corresponding to the number of pixel points in the target image. If the target bit is used as the target pixel value and the th layer of the homomorphic pixel value of its th layer of the sub-codeword, when encoding the th layer, use this sub-codeword to encode the target pixel value and the th layer of the homomorphic pixel value of its th layer, when decoding the encoding result of the th layer, decode the th layer of the homomorphic pixel value of the target pixel value into the target pixel value, then indicates the loss amount when the th homomorphic pixel value of the th layer of the target pixel value is decoded into the target pixel value, then reflects the average loss amount when all the pixel points corresponding to the homomorphic pixel values of the

[0046] 5. In response to the th layer loss degree being greater than the preset th layer loss threshold, increment the value of by one, and repeat steps 3 - 5 until the th layer loss degree is not greater than the preset th layer loss threshold, then stop the iteration.

[0047] 6. In response to the loss degree of the layer being not greater than a preset loss threshold of the layer, taking the target bit as the target pixel value and the homomorphic pixel value of its layer, and the sub-codeword of its layer, and denoting the length of the sub-codeword of the target pixel value and the homomorphic pixel value of its .

[0048] 7. Repeat steps 1 - 6 until all the sub-codewords of the layer of all candidate pixel values have been obtained and then stop the iteration.

[0049] 8. Increment the value of by one, and repeat steps 1 - 8 until there are no candidate pixel values in the layer and then stop the iteration.

[0050] It should be noted that the loss threshold of each layer determines the loss degree of each layer. The purpose of the present invention is for hierarchical compression transmission of remote sensing images. When the monitoring terminal receives the compression result, it can decompress and display the remote sensing image layer by layer. The decompression result of the first layer contains the global information in the corresponding channel of the remote sensing image. As the decompression layer is higher, the more details in the corresponding channel of the remote sensing image are contained in the decompression result. Therefore, the lower the layer, the higher the allowable loss degree, and the higher the layer, the lower the allowable loss degree.

[0051] Therefore, in this embodiment, the loss threshold of the first layer is 15, the loss threshold of the second layer is 10, the loss threshold of the third layer is 5, the loss threshold of the fourth layer is 1, and the loss threshold of the fifth layer is 0. It should be noted that when the loss threshold of the fifth layer is set to 0, the first five layers already contain all the information in the corresponding channel of the remote sensing image. Therefore, there will be no sixth layer, and there is no need to set the loss threshold of the sixth layer. In other embodiments, the implementer can set the loss threshold of each layer according to the actual implementation situation.

[0052] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, when obtaining the codewords of each pixel value by using the method corresponding to the first embodiment in step S003, the process of obtaining the sub-codewords of each layer of each pixel value is as follows: Taking 80 as the target pixel value, the codeword of 80 is 110, taking 1 as the target bit, the pixel values whose first 1 bit of the codeword is the same as 1 are 23, 24, 25, 79, 189, 192, then 23, 24, 25, 79, 189, 192 are the homomorphic pixel values of the first layer of 80. Then, when taking the target bit 1 as the first-layer sub-codeword of the target pixel value and its first-layer homomorphic pixel values, the first loss degree is (5×|23 - 80| + |24 - 80| + |25 - 80| + 3×|79 - 80| + 2×|189 - 80| + 3×|192 - 80|)÷(5 + 1 + 1 + 3 + 2 + 3)≈63.53. The first loss degree is greater than the loss threshold 15 of the first layer. Then, taking 11 as the target bit, the pixel values whose first 2 bits of the codeword are the same as 11 are 79, 192, then 79, 192 are the homomorphic pixel values of the first layer of 80. Then, when taking the target bit 11 as the first-layer sub-codeword of the target pixel value and its first-layer homomorphic pixel values, the first loss degree is , the first loss degree is greater than the loss threshold 15 of the first layer. Then, taking 110 as the target bit, there is no pixel value whose first 3 bits of the codeword are the same as 110. Then, 80 has no homomorphic pixel values in the first layer. Then, when taking the target bit 110 as the first-layer sub-codeword of the target pixel value and its first-layer homomorphic pixel values, the first loss degree is 0. The first loss degree is not greater than the loss threshold 15 of the first layer. Then, taking the target bit 110 as the first-layer sub-codeword of the pixel value 80. Similarly, for the sub-codewords of each layer of each pixel value, see Figure 8 .

[0053] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, when obtaining the codewords of each pixel value by using the method corresponding to the second embodiment in step S003, for the sub-codewords of each layer of each pixel value, see Figure 9 .

[0054] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, when obtaining the codewords of each pixel value using the method corresponding to the third embodiment in step S003, refer to the sub-codewords of each layer of each pixel value Figure 10 .

[0055] S004. Perform hierarchical encoding on each channel of the remote sensing image by using the sub-codewords of each layer of each pixel value.

[0056] Specifically, for each pixel point in the target image, encode the pixel point according to the codeword of the pixel value of the pixel point. During the encoding process, put the sub-codewords of each layer in the codeword into each layer respectively as the encoding result of the pixel point in each layer. Concatenate the encoding results of all pixel points in the same layer in sequence as the encoding result of the target image in this layer. Among them, if a certain pixel point does not have the sub-codeword of a certain layer, then skip this pixel point during the process of concatenating the encoding results of all pixel points in this layer.

[0057] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, if using Figure 8 the sub-codewords of each layer of each pixel value in to encode this sequence, the encoding result of the first layer is 110110010011101101110110110111001000001011000000000010110110100101101100110011000100001101010111111110111111011011, the encoding result of the second layer is 110101011100001, and there is no encoding result for the third layer.

[0058] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, if usingFigure 9 The sub - codewords of each layer for each pixel value in it encode this sequence. The encoding result of the first layer is 0000001010011000011000000011101011010110111011101111101000100000100100110000010011110111101100110001011111100110010101001010010, the encoding result of the second layer is 01, and there is no encoding result for the third layer.

[0059] Exemplarily, when the sequence formed by the pixel values of the pixel points in the target image is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, if Figure 10 the sub - codewords of each layer for each pixel value in it are used to encode this sequence, the encoding result of the first layer is 00000000000000000000011001001001000101010101010101100100111110111111110111101101, the encoding result of the second layer is 0011010011001000110101011010111001111, the encoding result of the third layer is 100010110011000, the encoding result of the fourth layer is 01, and there is no encoding result for the fifth layer.

[0060] It should be noted that Figure 8 、 Figure 9 、 Figure 10 the sub - codewords of each pixel value respectively correspond to the first embodiment, the second embodiment, and the third embodiment in step S003. Using Figure 8 、 Figure 9 、 Figure 10The sub - codeword pairs of each layer for each pixel value in the sequence {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 24, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190} are encoded. The lengths of the encoding results of the first layer are 114, 127, and 80 respectively. It can be seen that by obtaining the codewords of each pixel value according to the method in the third embodiment of step S003, then obtaining the sub - codewords of each layer for each pixel value, and encoding the sequence, the length of the encoding result of the first layer is the shortest. When transmitting the encoding result of the first layer to the monitoring terminal subsequently, the fewer bits need to be transmitted, and the transmission speed is faster, making the real - time monitoring of salt tides stronger.

[0061] S005: Transmit the encoding results of different layers to the monitoring terminal in sequence. The monitoring terminal decompresses the encoding results of each layer and performs real - time monitoring of salt tides according to the decompression results.

[0062] It should be noted that the encoding result of the first layer in the present invention can be decoded independently, and the decoding result contains the global information of the corresponding channel of the remote - sensing image and can be used to monitor salt tides. To ensure the real - time nature of salt - tide monitoring, the present invention transmits the encoding results of different layers in the hierarchical order of the encoding results.

[0063] In one embodiment, first transmit the encoding results of the first layer of each channel of the remote - sensing image to the monitoring terminal, and then transmit the encoding results of the second layer of each channel of the remote - sensing image, and so on until the sequential transmission of all hierarchical encoding results is completed.

[0064] In response to the monitoring terminal receiving the encoding results of the first layer of each channel of the remote - sensing image, decompress the encoding results of the first layer of each channel to obtain a decompressed image, and identify whether there is a salt - tide intrusion according to the decompressed image. In response to the monitoring terminal receiving the encoding results of the remaining layers of each channel of the remote - sensing image and there being a salt - tide intrusion, decompress the encoding results of the remaining layers in sequence, update the decompressed image according to the decompression results, and obtain the detailed information of the salt - tide intrusion according to the finally obtained decompressed image.

[0065] In another embodiment, transmit the encoding results of the first layer of each channel of the remote - sensing image to the monitoring terminal. In response to the monitoring terminal receiving the encoding results of the first layer of each channel of the remote - sensing image, decompress the encoding results of the first layer of each channel to obtain a decompressed image, and identify whether there is a salt - tide intrusion according to the decompressed image. In response to there being a salt - tide intrusion, transmit the encoding results of the remaining layers of each channel of the remote - sensing image to the monitoring terminal, decompress the encoding results of the remaining layers in sequence, update the decompressed image according to the decompression results, and obtain the detailed information of the salt - tide intrusion according to the finally obtained decompressed image.

[0066] Specifically, the specific method for decompressing the encoding results of the first layer of each channel is as follows: For the encoding result of the first layer of any channel, starting from the first undecoded bit in the encoding result, read the bits in the encoding result in the order from left to right. When there is a sub-codeword of the first layer of the pixel value that is the same as the read content, stop reading and use this pixel value as a decoding result. Repeat this process until all bits in the encoding result of the first layer of this channel have been decoded and stop the iteration. Fill all the decoding results into an empty matrix of the same size as this channel of the remote sensing image in order to obtain the decompression result of this channel. The decompression results of all channels form a decompressed image.

[0067] Exemplarily, when the encoding result of the first layer of a certain channel is 11011001001110110110110110111001000001011000000000010110110100101101100110011000100001101010111111110111111011011, and the sub-codewords of each layer of each pixel value are shown in Figure 8 At this time, starting from the first undecoded bit in the encoding result (i.e., the first bit of the encoding result), read the bits in the encoding result in the order from left to right. When 110 is read, the sub-codeword of the first layer of the pixel value 80 is the same as 110, so 80 is the first decoding result; then starting from the 4th bit in the encoding result, read the bits in the encoding result in the order from left to right. When 110 is read, the sub-codeword of the first layer of the pixel value 80 is the same as 110, so 80 is the second decoding result; then starting from the 7th bit in the encoding result, read the bits in the encoding result in the order from left to right. When 0100 is read, the sub-codeword of the first layer of the pixel value 78 is the same as 0100, so 78 is the second decoding result; and so on. All decoding results can be obtained, and the final sequence formed by all decoding results is: {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 22, 19, 22, 22, 22, 22, 22, 23, 23, 23, 22, 23, 23, 24, 24, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}.

[0068] Exemplarily, when the encoding result of the first layer of a certain channel is 0000001010011000011000000011101011010110111011101111101000100000100100110000010011110111101100110001011111100110010101001010010, the sub-codewords of each pixel value in each layer are referred to Figure 9 When, the sequence formed by all the final decoding results is: {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 77, 19, 20, 20, 21, 22, 22, 23, 23, 23, 22, 23, 23, 21, 25, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}.

[0069] Exemplarily, when the encoding result of the first layer of a certain channel is 00000000000000000000011001001001000101010101010101100100111110111111110111101101, the sub-codewords of each pixel value in each layer are referred to Figure 10 When, the sequence formed by all the final decoding results is: {80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 23, 20, 20, 20, 20, 23, 23, 23, 23, 23, 23, 23, 23, 20, 20, 192, 192, 190, 192, 192, 192, 190, 192, 190, 190}.

[0070] Furthermore, the specific method for sequentially decompressing the encoding results of the remaining layers is: Obtain all the decoding results in the layer that have the sub-codewords of the layer as the objects to be decoded, and sequentially decode all the objects to be decoded in order: For the current object to be decoded, use the sub-codewords of the layer that are the same as the sub-codewords of the current object to be decoded for all the pixel values as the homomorphic pixel values of the current object to be decoded. Starting from the first undecoded bit in the encoding result of the layer, read the bits in the encoding result of the layer in order from left to right. In response to the sub-codewords of the layer of the current object to be decoded being the same as the content read, stop reading, and use the current object to be decoded as the decoding result of the current object to be decoded in the layer; In response to any one of the homomorphic pixel values of the current object to be decoded layer The sub-codewords of the layer are the same as the read content, stop reading, and use this homomorphic pixel value as the decoding result of the current object to be decoded at the layer.

[0071] Replace each object to be decoded in the corresponding channel of the decompressed image with the decoding result of each object to be decoded at the layer, and implement the th update of the decompressed image.

[0072] Exemplarily, when the encoding result of the second layer of a certain channel is 110101011100001, and the sub-codewords of each pixel value in each layer are shown in Figure 8 , and the sequence formed by the decoding result of the first layer is {80, 80, 78, 79, 80, 79, 80, 80, 79, 78, 22, 19, 22, 22, 22, 22, 22, 23, 23, 23, 22, 23, 23, 24, 24, 189, 189, 190, 193, 192, 192, 190, 192, 190, 190}, if 22, 22, 22, 22, 22, 22, 22, 24, 24 in the decoding result of the first layer have the sub-codewords of the second layer, then the sequence formed by the objects to be decoded is {22, 22, 22, 22, 22, 22, 22, 24, 24}. For the first object to be decoded, 22, the sub-codewords of pixel values 20, 77, 21 in the first layer are the same as those of 22, so pixel values 20, 77, 21 are the homomorphic pixel values of the object to be decoded 22. Starting from the first undecoded bit in the encoding result (i.e., the first bit of the encoding result), read the bits in the encoding result in the order from left to right. When 110 is read, the sub-codewords of the homomorphic pixel value 77 in the first layer are the same as 110, so 77 is the decoding result of the first object to be decoded 22 at the second layer; for the second object to be decoded, 22, the homomorphic pixel values are 20, 77, 21. Starting from the 4th bit in the encoding result (i.e., the first bit of the encoding result), read the bits in the encoding result in the order from left to right. When 10 is read, the sub-codewords of the homomorphic pixel value 20 in the first layer are the same as 10, so 20 is the decoding result of the second object to be decoded 22 at the second layer; similarly, the decoding results of all objects to be decoded at the second layer can be obtained. When the encoding result of the first layer is 110110010011101101110110110111001000001011000000000010110110100101101100110011000100001101010111111110111111011011, and the encoding result of the second layer is 110101011100001, the decoding results of each layer are shown in Figure 11 .

[0073] Exemplarily, when the sub-codewords of each layer of each pixel value refer to Figure 9 , and the encoding result of the first layer is 0000001010011000011000000011101011010110111011101111101000100000100100110000010011110111101100110001011111100110010101001010010, and the encoding result of the second layer is 01, the decoding results of each layer refer to Figure 12 .

[0074] Exemplarily, when the sub-codewords of each layer of each pixel value refer to Figure 10 , and the encoding result of the first layer is 00000000000000000000011001001001000101010101010101100100111110111111110111101101, the encoding result of the second layer is 0011010011001000110101011010111001111, the encoding result of the third layer is 100010110011000, and the encoding result of the fourth layer is 01, the decoding results of each layer refer to Figure 13 .

[0075] It should be noted that the decompressed image is a remote sensing image. Identifying whether there is a salt tide intrusion based on the remote sensing image is a well-known technology. For example, a salt tide intrusion will cause changes in the color of the water body. Fresh water usually appears blue-green or dark blue, while the area affected by the salt tide intrusion may appear grayish-white or milky-yellow due to suspended salts, and a clear color boundary, i.e., a salt tide peak, may form at the confluence of fresh and salt water. Implementers can set identification rules based on this feature to identify whether there is a salt tide intrusion, which will not be elaborated in detail here.

[0076] It should be further noted that as the decompression layer gets higher, the more details in the corresponding channels of the remote sensing image are included in the decompression result, and the obtained decompressed image contains the complete information of the remote sensing image. Then, implementers can obtain the detailed information of the salt tide intrusion based on the finally obtained decompressed image.

[0077] The embodiment of the present invention also discloses a real-time salt tide monitoring system based on satellite remote sensing, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a real-time salt tide monitoring method based on satellite remote sensing according to the present invention.

[0078] The above system also includes other components well-known to those skilled in the art such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

Claims

1. A real-time monitoring method for salt tides based on satellite remote sensing, characterized in that, Comprising: Real-time collecting remote sensing images of a monitoring area; Using a variable-length coding algorithm to assign codewords to pixel values in each channel of the remote sensing image; Dividing the codewords of each pixel value into several layers according to the loss threshold of each layer to obtain sub-codewords of each layer of each pixel value; Using the sub-codewords of each layer of each pixel value to perform hierarchical coding on each channel of the remote sensing image, and sequentially transmitting the coding results of different layers to a monitoring terminal for real-time salt tide monitoring.

2. The real-time monitoring method of salt tide based on satellite remote sensing according to claim 1, characterized in that, The method for obtaining the sub-codewords of each layer of each pixel value is as follows: S101: The front of the code word bits are the same as the first bit of the codeword of the target pixel value The pixel value with the same bits is used as the target pixel value. Homomorphic pixel values ​​of the layer, The previous value of the target pixel The total length of the sub-codewords of the layer, For the S102: Determine the target bit as the target pixel value and its first The homomorphic pixel value of the layer The sub-codeword of the layer Loss degree, target bit is the codeword of target pixel value to bits; S103: In response to the The loss is not greater than the preset The loss threshold of the layer is taken as the target bit as the target pixel value and its The homomorphic pixel value of the layer The sub-codeword of the layer, otherwise, The value of is increased by 1, and steps S101-S103 are repeated until the target pixel value and its first pixel value are obtained. The homomorphic pixel value of the layer The iteration stops when the subcodeword of the layer is reached.

3. The real-time monitoring method of salt tide based on satellite remote sensing according to claim 2, characterized in that The said first The loss degree satisfies the expression: ； Among them, represents the loss degree when the target bit is used as the target pixel value and the sub-codeword of the layer of the homomorphic pixel value of the layer; loss degree, represents the number of homomorphic pixel values of the layer of the target pixel value; represents the th homomorphic pixel value of the layer of the target pixel value; represents the target pixel value; represents the number of pixel points corresponding to the th homomorphic pixel value of the layer of the target pixel value.

4. A real-time monitoring method for salt tides based on satellite remote sensing according to claim 1, characterized in that, The using the sub-codewords of each layer of each pixel value to perform hierarchical coding on each channel of the remote sensing image includes: For each pixel point in each channel of the remote sensing image, taking the sub-codewords of each layer corresponding to the pixel value of the pixel point as the coding results of the pixel point in each layer; concatenating the coding results of all pixel points in the same layer in sequence as the coding results of each channel of the remote sensing image in this layer.

5. A real-time monitoring method for salt tides based on satellite remote sensing according to any one of claims 1-4, characterized in that, The using a variable-length coding algorithm to assign codewords to pixel values in each channel of the remote sensing image includes: Obtaining the frequencies of each pixel value in the remote sensing image, and using Huffman coding to obtain the codewords of each pixel value according to the frequencies of each pixel value.

6. A real-time monitoring method for salt tides based on satellite remote sensing according to any one of claims 1-4, characterized in that, The using a variable-length coding algorithm to assign codewords to pixel values in each channel of the remote sensing image includes: Obtaining the frequencies of each pixel value in the remote sensing image, sorting the pixel values according to the frequencies of each pixel value, and using Shannon-Fano coding to obtain the codewords of each pixel value according to the sorted order of the pixel values and the frequencies of each pixel value.

7. A real-time monitoring method for salt tides based on satellite remote sensing according to claim 6, characterized in that The sorting the pixel values according to the frequencies of each pixel value includes: Sorting the pixel values in descending order according to the frequencies of each pixel value.

8. A real-time monitoring method for salt tides based on satellite remote sensing according to claim 6, characterized in that, The sorting the pixel values according to the frequencies of each pixel value includes: S201: Construct an empty sequence, denoted as the pixel sequence; S202: Add the pixel value with the highest frequency to the pixel sequence; S203: For all pixel values not yet added to the pixel sequence, determine the priority of the pixel value according to the frequency of the pixel value and the difference between the pixel value and the last pixel value in the current pixel sequence: , represents the priority of the -th pixel value not yet added to the pixel sequence; represents the -th pixel value not yet added to the pixel sequence; represents the last pixel value in the current pixel sequence; represents the -th pixel value not yet added to the pixel sequence; represents the absolute value symbol; Add the pixel value with the highest priority to the end of the pixel sequence; S204: Repeat S203 until all pixel values are added to the pixel sequence and stop the iteration.

9. A real-time monitoring method for salt tides based on satellite remote sensing according to claim 1, characterized in that, Also comprising: In response to the monitoring terminal receiving the coding result of the first layer of each channel of the remote sensing image, decompressing the coding result of the first layer of each channel to obtain a decompressed image, and identifying whether there is a salt tide intrusion according to the decompressed image; in response to the monitoring terminal receiving the coding results of the remaining layers of each channel of the remote sensing image and there being a salt tide intrusion, sequentially decompressing the coding results of the remaining layers, updating the decompressed image according to the decompression results, and obtaining detailed information on the salt tide intrusion according to the finally obtained decompressed image.

10. A real-time monitoring system for salt tides based on satellite remote sensing, characterized in that, Comprising: A processor and a memory, the memory storing computer program instructions, and when the computer program instructions are executed by the processor, implementing a real-time salt tide monitoring method based on satellite remote sensing according to any one of claims 1-9.

Images