Image transmission methods, systems, equipment and media for underwater wireless sensor networks

Abstract

This application provides an image transmission method, system, device, and medium for underwater wireless sensor networks. The method includes: acquiring an original image; using the original image as image blocks selected in the first screening, and further screening based on target conditions to process image blocks that do not meet the target conditions into N sub-blocks, and using these N sub-blocks as image blocks selected in the next screening for further screening, until a final screening is completed; acquiring the hierarchical information of each sub-block in the original image; encoding the sub-blocks that meet the target conditions to obtain encoded sub-blocks; and transmitting the encoded information stream to a central node based on the hierarchical information, in the order from the first layer to the last layer, using a trained deep learning joint source-channel model encoding. This application enables high-quality image transmission in underwater wireless sensor networks.

Description

Technical Field

[0001] This application relates to the field of underwater acoustic communication, and more specifically, to an image transmission method, system, device, and medium for an underwater wireless sensor network. Background Technology

[0002] To gain a comprehensive understanding of marine resources, we need to deploy observation equipment in the ocean to sense and monitor the various complex processes driving changes in the marine environment, and transmit the collected data to a data analysis center for processing. Especially during deep-sea resource development, in order to monitor deep-sea environmental information in real time, it is essential to transmit deep-sea images captured by underwater observation nodes equipped with cameras back to the base station in real time via acoustic signals.

[0003] In recent years, radio communication technology has been widely applied in daily life, becoming an important component of wireless sensor networks and greatly facilitating data collection. However, most wireless communication technologies rely on high-frequency electromagnetic waves, which suffer from rapid attenuation and limited communication distance both above and below sea level. In contrast, while underwater wired networks are stable and reliable, offering advantages such as high speed and low latency, their high cost and complex deployment and maintenance make large-scale deployment in vast ocean areas impractical. Although sound waves propagate quickly and travel long distances underwater, the narrow bandwidth and high power consumption of underwater acoustic communication make it difficult to support real-time transmission of high-quality images. Therefore, when transmitting images in underwater environments, compression and encoding techniques are needed to reduce data volume to adapt to bandwidth limitations and lower power consumption.

[0004] Therefore, this application provides an image transmission method, system, device, and medium for underwater wireless sensor networks to solve one of the aforementioned technical problems. Summary of the Invention

[0005] The purpose of this application is to provide an image transmission method, system, device and medium for underwater wireless sensor networks, which can solve at least one of the technical problems mentioned above.

[0006] The specific plan is as follows:

[0007] According to a specific embodiment of this application, in a first aspect, this application provides an image transmission method for an underwater wireless sensor network, applied to a detection node, comprising:

[0008] The process involves: acquiring the original image; using the original image as the image block selected in the first screening, and further screening based on target conditions to process image blocks that do not meet the target conditions into N sub-blocks, and using the N sub-blocks as the image blocks selected in the next screening for further screening, until the final screening is completed; wherein, the target conditions include that the sub-blocks meet the uniformity condition and that the size of the sub-blocks is less than a preset size; in the final screening, each sub-block satisfies at least one of the target conditions, and N is a positive integer greater than or equal to 4; acquiring the layer information of each sub-block in the original image; wherein, the layer information includes the layer number to which the sub-block belongs and the original image block to which the sub-block belongs; wherein, the layer number to which the sub-block belongs is based on the sub-block size, with the sub-block with the largest size corresponding to the first layer and the sub-block with the smallest size corresponding to the last layer; encoding the sub-blocks that meet the target conditions to obtain encoded sub-blocks; based on the layer information, transmitting the encoded information stream to the central node using a trained deep learning joint source-channel model encoding in the order from the first layer to the last layer; wherein, the encoded information stream is used to transmit the encoded sub-blocks layer by layer.

[0009] In one embodiment, encoding the sub-blocks that satisfy the target conditions to obtain encoded sub-blocks includes: data encoding the sub-blocks that satisfy the target conditions according to color values ​​or pixel values ​​to obtain data-encoded sub-blocks; wherein the data-encoded sub-blocks collectively represent the original image after data encoding; and re-encoding the original image after data encoding according to a joint source-channel coding scheme to obtain a re-encoded original image; wherein the re-encoded original image is collectively represented by the re-encoded sub-blocks; and the re-encoded sub-blocks are used as the encoded sub-blocks.

[0010] In one embodiment, the hierarchical information is based on a flag bit identifier; the method further includes: encoding and compressing the flag bit and the sub-block after data encoding based on a prefix encoding method or an arithmetic encoding method.

[0011] In one embodiment, the step of re-encoding the original image after data encoding according to the joint source-channel coding scheme to obtain the re-encoded original image includes: preprocessing the original image after data encoding to obtain a preprocessed original image; wherein the preprocessing includes scaling and / or normalization; and re-encoding the preprocessed original image according to the joint source-channel coding scheme based on deep learning to obtain the re-encoded original image.

[0012] In one embodiment, the step of transmitting the encoded information stream to the central node using trained deep learning joint source-channel model coding based on the hierarchical information and in the order from the first layer to the last layer includes: transmitting the encoded information stream to the central node using trained deep learning joint source-channel model coding based on the hierarchical information and in the order from the first layer to the last layer, and according to a preset limit number of layers; wherein the preset limit number of layers is used to limit the number of layers that can be transmitted.

[0013] According to a specific embodiment of this application, in a second aspect, this application provides an image transmission method for an underwater wireless sensor network, applied to a central node, comprising:

[0014] The system receives an encoded information stream transmitted by a probe node; wherein the encoded information stream includes encoded sub-blocks transmitted layer by layer by the probe node in the order from the first layer to the last layer; wherein each sub-block is used to jointly form the original image acquired by the probe node, and the layer number of the sub-block is characterized by the sub-block size, with the sub-block with the largest size corresponding to the first layer and the sub-block with the smallest size corresponding to the last layer; the system decodes the image blocks one by one according to the transmission order of the encoded information stream, and stitches the images according to the decoding order; the system monitors the image sharpness of the stitched image; and in response to detecting that the image sharpness of the stitched image is greater than or equal to a preset sharpness threshold, the system terminates the transmission of the encoded information stream.

[0015] In one embodiment, the step of decoding image blocks sequentially according to the transmission order of the encoded information stream includes: a joint source-channel decoding method based on deep learning, which decodes image blocks sequentially according to the transmission order of the encoded information stream.

[0016] According to a specific embodiment of this application, in a third aspect, this application provides an image transmission system for an underwater wireless sensor network, comprising:

[0017] A probe node is used to acquire the original image; select the image block in the first screening based on the original image, and further screen it based on target conditions to process the image block that does not meet the target conditions into N sub-blocks, and select the N sub-blocks as the image block for the next screening, and screen again until the last screening is completed, where N is a positive integer greater than or equal to 4; acquire the layer information of each sub-block in the original image; encode the sub-blocks that meet the target conditions to obtain encoded sub-blocks; and transmit the encoded information stream to the central node using a trained deep learning joint source-channel model encoding according to the layer information and in the order from the first layer to the last layer; the central node is used to receive the encoded information stream transmitted by the probe node; decode the image block sequentially according to the transmission order of the encoded information stream, and stitch the image according to the decoding order; monitor the image clarity of the stitched image; and terminate the transmission of the encoded information stream when the image clarity of the stitched image is detected to be greater than or equal to a preset clarity threshold.

[0018] According to a specific embodiment of this application, in a third aspect, this application provides an underwater acoustic device, including a memory, a processor, and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the method described in any one of the first aspects, or executes the computer program to implement the method described in any one of the second aspects.

[0019] According to a specific embodiment of this application, in a fourth aspect, this application provides a computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, when the computer program / instructions are executed by a processor, they implement the method described in any one of the first aspects, or implement the method described in any one of the second aspects.

[0020] Compared with the prior art, the above-described solutions of this application have at least the following beneficial effects:

[0021] The method provided in this application progressively segments the original image into sub-blocks that meet uniformity conditions and are smaller than a preset size, and obtains the layer information of each sub-block. Only sub-blocks that meet the conditions are encoded and transmitted. Because the image data is compressed into multiple sub-blocks, the amount of data transmitted can be significantly reduced, thereby improving transmission efficiency in bandwidth-constrained underwater environments. Specifically, the method of recursively segmenting the image and transmitting the encoded information stream layer by layer significantly reduces the amount of data transmitted. Attached Figure Description

[0022] Figure 1 A flowchart of an image transmission method for an underwater wireless sensor network is shown;

[0023] Figure 2 A flowchart illustrating the decomposition of a quadtree image is shown.

[0024] Figure 3 This illustrates a quadtree diagram after image decomposition;

[0025] Figure 4 A flowchart of an image transmission method for an underwater wireless sensor network according to an embodiment of this application is shown;

[0026] Figure 5 A flowchart of a method for re-encoding the original image after data encoding according to the joint source-channel coding scheme is shown;

[0027] Figure 6 A flowchart of another image transmission method for underwater wireless sensor networks is shown;

[0028] Figure 7 A flowchart of an image transmission method for an underwater wireless sensor network is shown;

[0029] Figure 8 A flowchart of an image transmission method for an underwater wireless sensor network is shown;

[0030] Figure 9 A schematic diagram of an image transmission system for an underwater wireless sensor network is shown.

[0031] Figure 10 The overall flowchart of image transmission in an underwater wireless sensor network scenario is shown.

[0032] Figure 11 A flowchart illustrating the information transmission process based on encoded data streams is shown.

[0033] Figure 12 The flowchart of the joint source-channel coding of this application is shown;

[0034] Figure 13 This is a block diagram of an underwater acoustic device for image transmission according to an exemplary embodiment. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0036] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the application. The singular forms “a,” “said,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0037] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0038] It should be understood that although the terms first, second, third, etc., may be used in the embodiments of this application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, first may also be referred to as second without departing from the scope of the embodiments of this application, and similarly, second may also be referred to as first.

[0039] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0040] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0041] It should be noted that any symbols and / or numbers present in the specification that are not marked in the accompanying drawings are not reference numerals.

[0042] This application is applied to underwater wireless detection scenarios, such as underwater wireless sensor networks. Detection nodes refer to nodes located underwater and equipped with cameras or other sensing devices. The main task of these nodes is to detect the environment of their surrounding sea area and transmit the captured images or other data to a central node in real time. The central node is the node that receives data from the detection nodes. The central node receives the image data transmitted from the detection nodes, decodes the received data, and presents the images for analysis.

[0043] The following embodiments of this application involve the description of image blocks and sub-blocks, where an image block is used to represent image data before block processing, and a sub-block is used to represent image data after block processing. In some embodiments, an image data can be regarded as both an image block and a sub-block. For example, an image block participating in the second screening can be regarded as one of the sub-blocks obtained after block processing in the first screening.

[0044] The optional embodiments of this application are described in detail below with reference to the accompanying drawings.

[0045] This application provides an embodiment of an image transmission method for an underwater wireless sensor network. The following is a description of that embodiment. Figure 1 The embodiments of this application will be described in detail.

[0046] Figure 1 A flowchart of an image transmission method for an underwater wireless sensor network is shown, such as... Figure 1 As shown, the method is applied to the probe node and includes the following steps S101 to S105.

[0047] Step S101: Obtain the original image.

[0048] The original image refers to the image initially acquired by the detection node without any processing, that is, the deep-sea image taken by the underwater observation node.

[0049] Step S102: The original image is used as the image block selected in the first screening, and the image block that does not meet the target condition is filtered into N sub-blocks. The N sub-blocks are then used as the image blocks selected in the next screening for further filtering, until the last screening is completed.

[0050] Where N is a positive integer greater than or equal to 4.

[0051] The target conditions include that the sub-blocks meet the uniformity condition and that the size of the sub-blocks is smaller than a preset size, which is predefined based on the underwater wireless sensor network.

[0052] For example, the following explanations assume N is 4, and during the final filtering, each sub-block satisfies at least one target condition.

[0053] Step S103: Obtain the layer information of each sub-block in the original image.

[0054] The layering information includes the layer number to which the sub-block belongs and the original image block to which the sub-block belongs.

[0055] The layer number to which a sub-block belongs is based on the sub-block size; the sub-block with the largest size corresponds to the first layer, and the sub-block with the smallest size corresponds to the last layer.

[0056] Step S104: Encode the sub-blocks that meet the target conditions to obtain the encoded sub-blocks.

[0057] Step S105: Based on the hierarchical information, the encoded information stream is transmitted to the central node using the trained deep learning joint source-channel model encoding in the order from the first layer to the last layer.

[0058] The encoded information stream is used to transmit the encoded sub-blocks layer by layer.

[0059] In this application, the uniformity condition refers to the standard used in image segmentation to determine whether an image patch can be considered to have the same or similar attributes (such as color, brightness, etc.). For example, when examining an image patch, if the pixel values ​​within the patch do not vary significantly, i.e., the color variance is less than a predefined threshold, then the patch is considered to meet the uniformity condition. If this condition is not met, the patch is further divided into multiple smaller patches, and the same check is performed on these smaller patches. This process continues until all sub-patterns meet the uniformity condition or reach a predetermined minimum patch size. Patches that meet the uniformity condition can directly encode their color or pixel values, while patches that do not meet the condition need to record their segmentation information and be further segmented. This strategy helps reduce data redundancy, thereby achieving image compression.

[0060] The method provided in this application progressively segments the original image into sub-blocks that meet uniformity conditions and are smaller than a preset size, and obtains the layer information of each sub-block. Only sub-blocks that meet the conditions are encoded and transmitted. Because the image data is compressed into multiple sub-blocks, the amount of data transmitted can be significantly reduced, thereby improving transmission efficiency in bandwidth-constrained underwater environments. Specifically, the method of recursively segmenting the image and transmitting the encoded information stream layer by layer significantly reduces the amount of data transmitted.

[0061] Figure 2 A flowchart illustrating the decomposition of a quadtree image is shown.

[0062] Figure 3A quadtree diagram after image decomposition is shown.

[0063] In the embodiments of this application, such as Figure 2 As shown, image patches that do not meet the target conditions are further subdivided into four smaller patches for more granular condition discrimination. Then, when each of the sub-patterned patches meets the target conditions, a filtering process is completed, resulting in multiple smaller patches that each meets at least one of the target conditions and has the same or different sizes. Further, as... Figure 3 As shown, the hierarchical information of each sub-block in the original image can be represented by a quadtree diagram. According to the branching direction shown in the diagram, it can be known which small-sized image blocks were obtained by dividing them from which large-sized image blocks, thus facilitating subsequent encoding and decoding.

[0064] In this application, the method of processing an image block into N sub-blocks can be characterized by a quadtree image decomposition process.

[0065] Figure 4 A flowchart illustrating an image transmission method for an underwater wireless sensor network according to an embodiment of this application is shown, such as... Figure 4 As shown, it includes the following sub-steps S401 to S402.

[0066] Sub-step S401: According to the color value or pixel value, each sub-block that meets the target condition is data encoded to obtain each sub-block after data encoding.

[0067] The encoded sub-blocks collectively represent the original image after data encoding.

[0068] Sub-step S402: The original image after data encoding is re-encoded according to the joint source-channel coding method to obtain the re-encoded original image.

[0069] The original image after re-encoding is represented by each re-encoded sub-block.

[0070] In this process, each sub-block after re-encoding is used as a sub-block after encoding.

[0071] In this embodiment, the encoding steps are further refined. A method is proposed to first encode the color or pixel values ​​of sub-blocks that meet the conditions, and then re-encode them using Joint Source Channel Coding (JSCC). This dual encoding mechanism not only further compresses the data but also better adapts to changes in channel conditions during transmission. By first encoding the color or pixel values, the amount of data can be reduced. Re-encoding using JSCC improves the robustness and stability of transmission. This method ensures that images can be accurately transmitted and recovered even under adverse channel conditions.

[0072] In one example of this application, the hierarchical information is based on flag identifiers.

[0073] In one embodiment of this application, the sub-blocks after encoding the flag bits and data can be encoded and compressed using prefix encoding or arithmetic encoding. This method can not only effectively reduce the space occupied by the flag bits, but also further compress the amount of data, thereby improving the efficiency of underwater acoustic transmission. At the same time, this method can manage image data more efficiently during transmission, reduce unnecessary information transmission, and further improve bandwidth utilization.

[0074] Figure 5 A flowchart illustrating a method for re-encoding the original image after data encoding according to a joint source-channel coding scheme is shown, as follows: Figure 5 As shown, it includes the following sub-steps S501 to S502.

[0075] Sub-step S501 involves preprocessing the original image after data encoding to obtain the preprocessed original image.

[0076] Preprocessing includes scaling and / or normalization.

[0077] Sub-step S502 involves re-encoding the preprocessed original image using a joint source-channel coding method based on deep learning, resulting in a re-encoded original image.

[0078] In this embodiment, the implementation of joint source-channel coding is improved by preprocessing the original image before encoding, ensuring the quality of the encoded image. The preprocessing step improves the statistical characteristics of the image, making it more suitable for subsequent encoding. This not only improves encoding efficiency but also enhances the quality of the decoded image. This method maintains high image quality during transmission while reducing data transmission volume, making it suitable for bandwidth-constrained underwater environments.

[0079] Figure 6 A flowchart of another image transmission method for underwater wireless sensor networks is shown, such as... Figure 6 As shown, the method is applied to the probe node and includes the following steps S601 to S605.

[0080] Step S601: Obtain the original image.

[0081] Step S602: The original image is used as the image block selected in the first filtering, and filtering is performed based on the target conditions to process the image block that does not meet the target conditions into N sub-blocks, and the N sub-blocks are used as the image blocks selected in the next filtering for filtering again, until the last filtering is completed.

[0082] The target conditions include that the sub-blocks meet the uniformity condition and that the size of the sub-blocks is smaller than the preset size.

[0083] For example, during the last filtering, each sub-block satisfies at least one target condition.

[0084] Step S603: Obtain the layer information of each sub-block in the original image. The layer information includes the layer number of the sub-block and the original image block to which the sub-block belongs.

[0085] The layer number to which a sub-block belongs is based on the sub-block size; the sub-block with the largest size corresponds to the first layer, and the sub-block with the smallest size corresponds to the last layer.

[0086] Step S604: Encode the sub-blocks that meet the target conditions to obtain the encoded sub-blocks.

[0087] Step S605: Based on the hierarchical information, the encoded information stream is transmitted to the central node in the order from the first layer to the last layer and according to the preset limit number of layers, using a trained deep learning joint source-channel model.

[0088] The preset limit on the number of layers is used to limit the number of layers that can be transmitted.

[0089] This application also provides another image transmission method for underwater wireless sensor networks, applied to a central node, for functional cooperation with the detection nodes mentioned in the above embodiments to complete image transmission. The interpretation of the same names is the same as in the above embodiments, and it has the same technical effects as in the above embodiments, so it will not be repeated here.

[0090] Figure 7 A flowchart of an image transmission method for an underwater wireless sensor network is shown, such as... Figure 7 As shown, the method is applied to the central node and includes the following steps S701 to S704.

[0091] Step S701: Receive the encoded information stream transmitted by the probe node.

[0092] The encoded information stream includes each encoded sub-block transmitted by the probe node in the order from the first layer to the last layer.

[0093] Each sub-block is used to jointly form the original image collected by the detection node. The layer to which a sub-block belongs is represented by its size. The sub-block with the largest size corresponds to the first layer, and the sub-block with the smallest size corresponds to the last layer.

[0094] Step S702: Decode the image blocks one by one according to the transmission order of the encoded information stream, and stitch the images together according to the decoding order.

[0095] Step S703: Monitor the image clarity after image stitching.

[0096] Step S704: In response to detecting that the image clarity after image stitching is greater than or equal to a preset clarity threshold, the transmission of the encoded information stream is terminated.

[0097] Figure 8 A flowchart of an image transmission method for an underwater wireless sensor network is shown, such as... Figure 8 As shown, the method is applied to the central node and includes the following steps S801 to S805.

[0098] Step S801: Receive the encoded information stream transmitted by the probe node.

[0099] The encoded information stream includes each encoded sub-block transmitted by the probe node in the order from the first layer to the last layer.

[0100] Each sub-block is used to jointly form the original image collected by the detection node. The layer to which a sub-block belongs is represented by its size. The sub-block with the largest size corresponds to the first layer, and the sub-block with the smallest size corresponds to the last layer.

[0101] Step S802: Based on the deep learning-based joint source-channel decoding method, image blocks are decoded sequentially according to the transmission order of the encoded information stream, and the images are stitched together according to the decoding order.

[0102] Step S803: Monitor the image clarity after image stitching.

[0103] Step S804: In response to detecting that the image clarity after image stitching is greater than or equal to a preset clarity threshold, the transmission of the encoded information stream is terminated.

[0104] This application also provides system embodiments that follow the above embodiments, for implementing the method steps of the above embodiments. The interpretation of the same names is the same as that of the above embodiments, and they have the same technical effects as those of the above embodiments, so they will not be repeated here.

[0105] Figure 9 A schematic diagram of an image transmission system for an underwater wireless sensor network is shown, such as... Figure 9 As shown, system 900 includes a detection node 901 and a central node 902.

[0106] Detector node 901 is used to acquire the original image. The original image is used as the image patch selected in the first filtering stage, and further filtering is performed based on target conditions. Image patches that do not meet the target conditions are processed into N sub-blocks, and these N sub-blocks are used as the image patches selected in the next filtering stage for further filtering, until the final filtering is completed. The hierarchical information of each sub-block in the original image is acquired. Sub-blocks that meet the target conditions are encoded to obtain encoded sub-blocks. The encoded information stream is then transmitted to the central node using a trained deep learning joint source-channel model, encoded according to the hierarchical information and in the order from the first layer to the last layer.

[0107] The central node 902 receives the encoded information stream transmitted by the probe nodes. It decodes image blocks sequentially according to the transmission order of the encoded information stream and then stitches the images together in the same order. It monitors the sharpness of the stitched image and terminates the transmission of the encoded information stream if the sharpness of the stitched image is greater than or equal to a preset sharpness threshold.

[0108] Figure 10 The overall flowchart of image transmission in an underwater wireless sensor network scenario is shown.

[0109] For example, such as Figure 10 As shown, x represents each sub-block with a quadtree structure, and y represents the original image after joint source-channel coding. This represents the original image after joint source-channel coding, output following transmission via the physical channel. This represents the original image reconstructed by the central node. First, the original image is recursively divided into multiple sub-blocks until each sub-block satisfies a uniformity condition or reaches a predetermined minimum block size. The hierarchical information of these sub-blocks is recorded, forming a quadtree structure. Next, these sub-blocks are encoded using Joint Source Channel Coding (JSCC) to adapt to the bandwidth-constrained underwater transmission environment. The encoded image data is transmitted to the central node via a physical channel. During transmission, the JSCC encoder converts the image data into a signal suitable for transmission through the physical channel. After receiving the signal, the central node demodulates the signal through the physical channel to recover the encoded data, and then uses the JSCC decoder to decode the data, ultimately recovering the original image. The entire process is implemented through hierarchical transmission, ensuring efficient and high-quality transmission of image data under bandwidth-constrained conditions.

[0110] Figure 11 A flowchart of information transmission based on encoded data streams is shown.

[0111] For example, such as Figure 11As shown, the probe node recursively divides the original image into multiple sub-blocks using quadtree hierarchical encoding, and encodes the sub-blocks that satisfy the uniformity condition. The encoded image information is transmitted layer by layer to the central node in order from the first layer to the last layer. After transmitting information of the depth layer (in this example, depth represents the layer number), the central node evaluates the sharpness of the received image. If the central node determines that the image is sufficiently sharp, it sends a stop transmission command to the probe node, which then stops transmitting further information at deeper levels. In this way, the system can effectively utilize bandwidth resources while ensuring image quality.

[0112] Figure 12 A flowchart of the joint source-channel coding process of this application is shown.

[0113] For example, such as Figure 12 As shown, x represents each sub-block with a quadtree structure, and y represents the original image after joint source-channel coding. This represents the original image after joint source-channel coding, output following transmission via the physical channel. This represents the original image reconstructed by the central node. At the transmitting end (i.e., the probe node), the original image is first preprocessed, including scaling and normalization, to meet the encoder's input requirements. The preprocessed image data is then fed into the joint source-channel encoder for efficient encoding. The encoded data is transmitted through a physical channel, which may contain noise and interference, affecting signal quality. Upon receiving the signal, the receiving end (i.e., the central node) first demodulates the signal to recover the encoded data. Then, the joint source-channel decoder decodes the data to recover the original image. Finally, post-processing, such as denoising and sharpening, is performed on the received image to improve image quality. This entire process ensures efficient and high-quality transmission of image data even under bandwidth-constrained and interference-prone conditions.

[0114] In this application, considering the bandwidth limitations of underwater acoustic communication, an image transmission coding algorithm for underwater wireless sensor networks is proposed. This algorithm integrates the advantages of progressive image coding and joint source-channel image transmission algorithms. The design comprehensively considers the complexity of the image and the variability of the underwater environment. Specifically, the overall process first uses quadtree hierarchical coding technology to process the original image, generating an image of a specific level. Then, it is compressed and encoded using a JSCC encoder. Finally, the encoded data is transmitted to the receiving end, where the JSCC decoder decodes and reconstructs the image.

[0115] In this application, the progressive image coding algorithm focuses on processing the color information of each pixel in the image. By employing image block-based techniques, it utilizes the similarity and differences of local regions in the image to segment it into increasingly smaller square blocks. Quadtree hierarchical coding is an effective compression method that recursively divides the image into N sub-blocks until each sub-block satisfies a uniformity condition or reaches a predetermined minimum block size, thereby reducing data redundancy and achieving efficient compression. This process includes four basic steps: image segmentation, recursive segmentation, encoding, and decoding. It aims to transmit image information progressively in a hierarchical manner using a depth-first traversal approach, ensuring that the central node can decide whether to continue receiving deeper layers of data based on the quality of the received information.

[0116] Correspondingly, joint source-channel coding (JSCC) algorithms aim to maximize the performance of communication systems by jointly designing source coding and channel coding. In the field of image transmission, JSCC technology can effectively combine the statistical characteristics of the source and the transmission characteristics of the channel to optimize image quality and efficiency during transmission. In particular, it can overcome the "cliff effect" in traditional separate source coding and channel coding systems, i.e., the performance drops sharply when channel conditions are below a certain threshold. Based on this, this application introduces a deep learning-based JSCC image transmission scheme, which includes three parts: a transmitter, a physical channel, and a receiver, realizing the entire process from image preprocessing to final decoding output. The transmitter encodes the image using a deep learning-based encoder, performing joint source-channel coding considering channel characteristics. The receiver uses a deep learning-based decoder to decode the encoded data, recover the original image, and improve image quality through post-processing. This scheme not only improves the quality and efficiency of image transmission but also enhances the system's ability to resist channel noise and interference.

[0117] Figure 13 This is a block diagram of an underwater acoustic device 1300 for image transmission according to an exemplary embodiment.

[0118] like Figure 13As shown, one embodiment of this application provides an underwater acoustic device 1300. The underwater acoustic device 1300 includes a memory 1301, a processor 1302, and an input / output (I / O) interface 1303. The memory 1301 is used to store instructions. The processor 1302 is used to execute the image transmission method of the underwater wireless sensor network according to the embodiments of this application by calling the instructions stored in the memory 1301. The processor 1302 is connected to both the memory 1301 and the I / O interface 1303, for example, via a bus system and / or other forms of connection mechanisms (not shown). The memory 1301 can be used to store programs and data, including the program for the image transmission method of the underwater wireless sensor network involved in the embodiments of this application. The processor 1302 executes various functional applications and data processing of the underwater acoustic device 1300 by running the program stored in the memory 1301.

[0119] In this embodiment, the processor 1302 can be implemented using at least one of the following hardware forms: digital signal processor (DSP), field programmable gate array (FPGA), and programmable logic array (PLA). The processor 1302 can be one or a combination of several of the following: central processing unit (CPU) or other processing units with data processing capability and / or instruction execution capability.

[0120] The memory 1301 in this embodiment may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0121] In this embodiment, the I / O interface 1303 can be used to receive input instructions (such as numerical or character information, and to generate key signal inputs related to user settings and function control of the underwater acoustic device 1300), and can also output various information (such as images or sounds) to the outside. In this embodiment, the I / O interface 1303 may include one or more of the following: a physical keyboard, function keys (such as volume control keys, power buttons, etc.), a mouse, a joystick, a trackball, a microphone, a speaker, and a touch panel.

[0122] In some embodiments, this application provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, perform any of the methods described above.

[0123] In some embodiments, this application provides a computer program product comprising a computer program that, when executed by a processor, performs any of the methods described above.

[0124] Although the operations are described in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order or serial order shown, or requiring all of the operations shown to obtain the desired result. In certain environments, multitasking and parallel processing may be advantageous.

[0125] The methods and apparatus of this application can be implemented using standard programming techniques, utilizing rule-based logic or other logic to implement various method steps. It should also be noted that the terms "apparatus" and "module" as used herein and in the claims are intended to include implementations using one or more lines of software code and / or hardware implementations and / or devices for receiving input.

[0126] Any step, operation, or procedure described herein may be performed or implemented using one or more hardware or software modules, either alone or in combination with other devices. In one embodiment, the software module is implemented using a computer program product comprising a computer-readable medium containing computer program code, which is executable by a computer processor to perform any or all of the described steps, operations, or procedures.

[0127] The foregoing description of implementations of this application has been provided for illustrative and descriptive purposes. The foregoing description is not exhaustive and is not intended to limit this application to the exact forms disclosed. Various modifications and variations may exist in accordance with the foregoing teachings, or may arise from practice of this application. These embodiments were chosen and described to illustrate the principles of this application and its practical application, enabling those skilled in the art to utilize this application in various implementations and modifications to suit the specific purpose of the concept.

[0128] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0129] It can be further understood that, unless otherwise specified, "connection" includes both direct connections where no other components exist between the two parties and indirect connections where other components exist between them.

[0130] It is further understood that although the operations are described in a specific order in the accompanying drawings in the embodiments of this application, this should not be construed as requiring these operations to be performed in the specific order or serial order shown, or requiring all the operations shown to be performed to obtain the desired result. In certain environments, multitasking and parallel processing may be advantageous.

[0131] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the field of this application that are not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0132] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

[0133] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An image transmission method for an underwater wireless sensor network, characterized in that, Applied to a detection node, the method includes: Obtain the original image; The original image is used as the image block selected in the first filtering, and further filtering is performed based on target conditions. Image blocks that do not meet the target conditions are processed into N sub-blocks, and the N sub-blocks are used as the image blocks selected in the next filtering for further filtering, until the final filtering is completed. The target conditions include that the sub-blocks meet the uniformity condition and that the size of the sub-blocks is less than a preset size. In the final filtering, each sub-block meets at least one of the target conditions, and N is a positive integer greater than or equal to 4. Obtain the layering information of each sub-block in the original image; wherein, the layering information includes the layer number to which the sub-block belongs and the original image block to which the sub-block belongs; wherein, the layer number to which the sub-block belongs is based on the sub-block size, with the sub-block with the largest size corresponding to the first layer and the sub-block with the smallest size corresponding to the last layer; Encode the sub-blocks that meet the target conditions to obtain the encoded sub-blocks; Based on the hierarchical information, the encoded information stream is transmitted to the central node using a trained deep learning joint source-channel model encoding, in the order from the first layer to the last layer; wherein the encoded information stream is used to transmit the encoded sub-blocks layer by layer. The process of encoding sub-blocks that satisfy the target conditions to obtain encoded sub-blocks includes: Data encoding is performed on each sub-block that meets the target condition according to color value or pixel value to obtain each sub-block after data encoding; wherein, each sub-block after data encoding collectively represents the original image after data encoding. The original image after data encoding is re-encoded according to the joint source-channel coding method to obtain the re-encoded original image; wherein, the re-encoded original image is jointly represented by each re-encoded sub-block; Each sub-block after re-encoding is used as the encoded sub-block; The step of transmitting the encoded information stream to the central node using a trained deep learning joint source-channel model encoding based on the hierarchical information and in the order from the first layer to the last layer includes: Based on the hierarchical information, the encoded information stream is transmitted to the central node in the order from the first layer to the last layer and according to a preset limit of the number of layers, using a trained deep learning joint source-channel model. The preset limit on the number of layers is used to limit the number of layers that can be transmitted.

2. The method according to claim 1, characterized in that, The hierarchical information is based on flag bits; The method further includes: The flag bits and the sub-blocks after data encoding are encoded and compressed based on either prefix encoding or arithmetic encoding.

3. The method according to claim 1, characterized in that, The step of re-encoding the original image after data encoding according to the joint source-channel coding method to obtain the re-encoded original image includes: The original image after data encoding is preprocessed to obtain a preprocessed original image; wherein, the preprocessing includes scaling and / or normalization. The preprocessed original image is re-encoded using a deep learning-based joint source-channel coding method to obtain a re-encoded original image.

4. An image transmission method for an underwater wireless sensor network, characterized in that, Applied to the central node, the method includes: Receive the encoded information stream transmitted by the probe node as described in claim 1; wherein the encoded information stream includes each encoded sub-block transmitted by the probe node layer by layer in the order from the first layer to the last layer; wherein each sub-block is used to jointly form the original image acquired by the probe node, and the layer number to which the sub-block belongs is characterized by the sub-block size, with the sub-block with the largest size corresponding to the first layer and the sub-block with the smallest size corresponding to the last layer. Image blocks are decoded sequentially according to the transmission order of the encoded information stream, and images are stitched together according to the decoding order. Monitor the image clarity after image stitching; In response to the detection that the image clarity after image stitching is greater than or equal to a preset clarity threshold, the transmission of the encoded information stream is terminated.

5. The method according to claim 4, characterized in that, The step of decoding image blocks sequentially according to the transmission order of the encoded information stream includes: The deep learning-based joint source-channel decoding method decodes image blocks sequentially according to the transmission order of the encoded information stream.

6. An image transmission system for an underwater wireless sensor network, characterized in that, The system includes: A probe node is used to acquire the original image; the original image is used as the image block selected in the first screening, and further screening is performed based on target conditions to process the image block that does not meet the target conditions into N sub-blocks, and the N sub-blocks are used as the image blocks selected in the next screening for further screening, until the last screening is completed, where N is a positive integer greater than or equal to 4; the layer information of each sub-block in the original image is acquired; the sub-blocks that meet the target conditions are encoded to obtain the encoded sub-blocks; and the encoded information stream is transmitted to the central node using a trained deep learning joint source-channel model encoding based on the layer information, in the order from the first layer to the last layer. The process of encoding sub-blocks that satisfy the target conditions to obtain encoded sub-blocks includes: Data encoding is performed on each sub-block that meets the target condition according to color value or pixel value to obtain each sub-block after data encoding; wherein, each sub-block after data encoding collectively represents the original image after data encoding. The original image after data encoding is re-encoded according to the joint source-channel coding method to obtain the re-encoded original image; wherein, the re-encoded original image is jointly represented by each re-encoded sub-block; Each sub-block after re-encoding is used as the encoded sub-block; The step of transmitting the encoded information stream to the central node using a trained deep learning joint source-channel model encoding based on the hierarchical information and in the order from the first layer to the last layer includes: Based on the hierarchical information, the encoded information stream is transmitted to the central node in the order from the first layer to the last layer and according to a preset limit of the number of layers, using a trained deep learning joint source-channel model. The preset layer limit is used to limit the number of layers that can be transmitted. The central node is used to receive the encoded information stream transmitted by the probe node; decode image blocks sequentially according to the transmission order of the encoded information stream, and stitch the images together according to the decoding order; monitor the image clarity after image stitching; and terminate the transmission of the encoded information stream when the image clarity after image stitching is detected to be greater than or equal to a preset clarity threshold.

7. An underwater acoustic device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method of any one of claims 1-3, and / or executes the computer program to implement the method of any one of claims 4-5.

8. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When executed by a processor, the computer program / instruction implements the method described in any one of claims 1-3, and / or implements the method described in any one of claims 4-5.

Images