Emergency communication transmission method and system based on Beidou satellite

Abstract

The present invention belongs to the field of communication transmission technology, and specifically relates to an emergency communication transmission method and system based on Beidou satellites, the method comprising: dividing an image to be transmitted in emergency communication into image blocks of the same size, forming a key image with the grayscale values ​​of the center points of all image blocks, converting a binary number composed of the th mark values ​​of all neighborhood points in each image block into a decimal number, and using the th representative grayscale value of the image block as the th representative grayscale value of the image block, forming a th supplementary image with the th representative grayscale value of all image blocks, performing emergency communication transmission on the key image and all supplementary images, wherein the first mark value of the neighborhood point is obtained according to the grayscale value size relationship between the neighborhood point and the center point, and the th mark value of the neighborhood point is obtained according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs. The present invention improves the effectiveness and efficiency of communication transmission by compressing the image to be transmitted.

Description

Technical Field

[0001] The present invention relates to the field of communication transmission technology, and more specifically, to a Beidou satellite-based emergency communication transmission method and system. Background Art

[0002] Smart devices rely on 4G or 5G base station wireless networks. When they exceed the coverage of the base station network, smart devices will lose their communication function. In sparsely populated areas, there is usually no 4G or 5G network coverage. At this time, satellite communication is particularly important. With the development of Beidou positioning and navigation technology, Beidou's first, second, and third generation navigation and communication systems have been built, and Beidou communication technology is also constantly developing into the field of civil technology. Engineering construction in some remote areas is inevitable. In areas without 4G or 5G network coverage, it is very important to ensure convenient communication and safety of construction workers.

[0003] In the related technology, for example, the Chinese patent application document with application publication number CN117692047A discloses a mobile monitoring system based on ad hoc satellite communication, including: a mobile monitoring drone subsystem, which is used to use a drone cluster equipped with a satellite terminal to collect cluster images, and transmit the cluster image collection results back to the emergency monitoring platform through the ad hoc satellite subsystem; the ad hoc satellite subsystem is used to perform satellite ad hoc networking and provide a satellite network to transmit the cluster image collection results to the emergency monitoring platform; the emergency monitoring platform is used to receive the cluster image collection results through the ad hoc satellite subsystem and display them on the screen; the invention also provides a satellite ad hoc network and provides a method for transmitting the cluster image collection results to the emergency monitoring platform through a satellite network. The invention solves the problems of unstable satellite communication and waste of communication resources due to clock synchronization in existing emergency communications.

[0004] In the process of emergency communication, in addition to communication transmission stability and communication resources, communication transmission effectiveness and communication transmission efficiency also need to be considered. High-resolution images and videos will seriously affect communication transmission effectiveness and communication transmission efficiency, so images or videos need to be compressed and transmitted. Summary of the invention

[0005] In order to solve the above-mentioned technical problem that in the process of emergency communication, high-resolution images and videos will seriously affect the effectiveness and efficiency of communication transmission, and therefore it is necessary to compress and transmit the images or videos, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides an emergency communication transmission method based on Beidou satellite, comprising: dividing an image to be transmitted in emergency communication into image blocks of the same size, wherein the image blocks include a central point and a plurality of neighborhood points; forming a key image from the central points of all the image blocks; and combining the first and second neighborhood points of all the neighborhood points in each image block. The binary number composed of the tag value is converted into a decimal number as the first Represents the gray value, and the Represents the gray value composition Supplementary images, , is a preset value; emergency communication transmission is performed on the key image and multiple supplementary images; wherein the first mark value of the neighborhood point The method of obtaining is: if the gray value of the neighborhood point is greater than the gray value of the center point, Equal to the first digit, otherwise =Equal to the second digital number; performing multi-peak Gaussian fitting on the grayscale differences of all neighborhood points in the image, and using the obtained Gaussian mixture model as the Gaussian mixture model to which the first mark value of the neighborhood point belongs; , the neighboring point Tagged Value The method of obtaining is: if the grayscale difference of the neighborhood points belongs to the sub-mixture model on the left, is equal to the first code, and the sub-mixture model on the left is used as the first The Gaussian mixture model to which the tag value belongs ,otherwise is equal to the second digit, and the sub-mixing model on the right is taken as ; The grayscale difference of the neighborhood points refers to the absolute value of the difference between the grayscale values ​​of the neighborhood points and the center point. The sub-mixture model is based on the segmentation threshold Gaussian Mixture Model Obtained by segmentation, is the neighborhood point The Gaussian mixture model to which the tag value belongs, Equal composition The valley between the middle two Gaussian components of all Gaussian components.

[0007] The present invention forms a key image with the center points of all image blocks in the image to be transmitted. The obtained key image is equivalent to a downsampled image of the image to be transmitted. The number of pixels of the image is directly reduced through downsampling. The reduction in the number of pixels means a reduction in the amount of data to be transmitted, thereby achieving compression of the image to be transmitted, thereby improving the effectiveness and efficiency of communication transmission.

[0008] After acquiring the key image, the present invention first acquires the first marking value of the neighborhood point through the grayscale value size relationship between the neighborhood point and the center point, and then acquires the first supplementary image, and then acquires each marking value of the neighborhood point according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs, and then acquires other supplementary images, thereby continuously narrowing the range to which the grayscale difference between the neighborhood point and the center point belongs. After transmitting the supplementary images, the receiver can obtain the grayscale value size relationship between the neighborhood point and the center point and the range to which the grayscale difference between the neighborhood point and the center point belongs according to each supplementary image, and then when upsampling the key image to restore the transmitted image, the grayscale value of each neighborhood point after upsampling can be limited; compared with directly restoring the transmitted image by upsampling the key image, more image features are retained and the image quality is better, which is helpful for subsequent image analysis, target recognition and other operations.

[0009] Preferably, the images to be transmitted in the emergency communication include static images and I-frame images in the video.

[0010] Preferably, the size of the image block is ,in, for An integer in the range.

[0011] Preferably, the grayscale differences of all neighborhood points in the image are subjected to multi-peak Gaussian fitting, and the obtained Gaussian mixture model is used as the Gaussian mixture model to which the first label value of the neighborhood point belongs, including: constructing a grayscale difference histogram according to the grayscale differences of all neighborhood points in the image; performing multi-peak Gaussian fitting on the grayscale difference histogram to obtain a Gaussian mixture model , It is composed of multiple Gaussian components, each of which has a peak, and there is a trough between two adjacent Gaussian components; Gaussian Components The fitting error and the difference between the predicted frequency at the peak and the predicted frequency at the trough are calculated. Gaussian Components The probability of retention is the highest. Gaussian mixture model composed of Gaussian components as the Gaussian mixture model to which the first label value of the neighborhood point belongs ;in, .

[0012] Preferably, the horizontal axis of the grayscale difference histogram is the grayscale differences, and the vertical axis is the frequency of the grayscale differences; the frequency of the grayscale difference is equal to the ratio of the number of all neighborhood points corresponding to the grayscale difference to all neighborhood points in the image.

[0013] Preferably, the Gaussian Components The retention probability satisfies the expression: ; In the formula, For the Gaussian Components The probability of retention, For the Gaussian Components The fitting error, Gaussian component The predicted frequency at the peak of It is Gaussian Components With Gaussian Components The predicted frequency of the trough between It is Gaussian Components With Gaussian Components The predicted frequency at the trough between .

[0014] For each Gaussian component, the difference between the predicted frequency at the peak and the predicted frequency at the trough is calculated, which reflects the model's ability to capture the histogram features. The fitting error reflects the accuracy of the model in capturing the histogram features. According to the fitting error and probability difference of each Gaussian component, the retention probability of each Gaussian component is calculated, so that the Gaussian component determined subsequently according to the retention probability can accurately reflect the grayscale differences of all neighborhood points in the image.

[0015] Preferably, the Gaussian Components The fitting error satisfies the expression: ; In the formula, For the Gaussian Components The fitting error, For the Gaussian Components The number of all grayscale differences involved, For the Gaussian Components The first The frequency of grayscale differences, For the Gaussian Components The first The grayscale difference is Gaussian Components The predicted frequency of is the absolute value function.

[0016] Preferably, when the key image and the plurality of supplementary images are transmitted for emergency communication, the key image, the first supplementary image, the second supplementary image, ..., the Supplementary images, The complementary images are transmitted in order.

[0017] According to the key image, the first supplementary image, the second supplementary image, ..., Supplementary images, The supplementary images are transmitted in sequence. When the situation is urgent, the image can be restored directly based on the currently received key image and multiple supplementary images, or even directly based on the key image, thereby improving the communication transmission efficiency.

[0018] Preferably, when the key image and the plurality of supplementary images are transmitted for emergency communication, each time a key image or a supplementary image is acquired, the acquired key image or supplementary image is transmitted.

[0019] Since images can be restored based on the key image and different numbers of supplementary images, the image features retained by the final restored image are different and the image quality is different; therefore, when the key image and the supplementary images are transmitted for emergency communication, it is possible to obtain the key image and the supplementary images while transmitting them, thereby improving the communication transmission efficiency.

[0020] In a second aspect, the present invention provides an emergency communication transmission system based on Beidou satellite, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned emergency communication transmission method based on Beidou satellite is implemented.

[0021] By adopting the above technical solution, the above-mentioned Beidou satellite-based emergency communication transmission method is generated into a computer program and stored in a memory to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0022] The beneficial effects of the present invention are:

[0023] The present invention forms a key image from the center points of all image blocks in the image to be transmitted. The obtained key image is equivalent to a downsampled image of the image to be transmitted. The number of pixels of the image is directly reduced by downsampling. The reduction in the number of pixels means that the amount of data to be transmitted is reduced, which realizes the compression of the image to be transmitted, thereby improving the effectiveness and efficiency of communication transmission.

[0024] Furthermore, after acquiring the key image, the present invention first acquires the first marking value of the neighborhood point through the grayscale value size relationship between the neighborhood point and the center point, and then acquires the first supplementary image, and then acquires each marking value of the neighborhood point according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs, and then acquires other supplementary images, thereby continuously narrowing the range to which the grayscale difference between the neighborhood point and the center point belongs. After transmitting the supplementary images, the receiver can obtain the grayscale value size relationship between the neighborhood point and the center point and the range to which the grayscale difference between the neighborhood point and the center point belongs according to each supplementary image, and then when upsampling the key image to restore the transmitted image, the grayscale value of each neighborhood point after upsampling can be limited; compared with directly restoring the transmitted image by upsampling the key image, more image features are retained and the image quality is better, which is helpful for subsequent image analysis, target recognition and other operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below through the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0026] Figure 1 is a flow chart schematically illustrating an emergency communication transmission method based on Beidou satellite in the present invention;

[0027] Figure 2 is a schematic diagram showing an image to be transmitted;

[0028] Figure 3 Schematically shows image blocks of different sizes;

[0029] Figure 4 It is schematically shown Figure 2 The key image of the image to be transmitted is shown;

[0030] Figure 5 It is schematically shown Figure 2 a first complementary image of the image to be transmitted;

[0031] Figure 6 is a schematic diagram showing the Gaussian mixture model obtained by multi-peak Gaussian fitting ;

[0032] Figure 7 It is schematically shown Figure 2 Shown is a second complementary image of the image to be transmitted. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0034] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0035] The embodiment of the present invention discloses an emergency communication transmission method based on Beidou satellite, referring to Figure 1 , including steps S1 to S4:

[0036] S1. Divide an image to be transmitted in emergency communication into multiple image blocks.

[0037] It should be noted that smart devices rely on 4G or 5G base station wireless networks. When they exceed the coverage of the base station network, smart devices will lose their communication function. In sparsely populated areas, there is usually no 4G or 5G network coverage. At this time, satellite communication is particularly important. With the development of Beidou positioning and navigation technology, Beidou's first, second, and third generation navigation and communication systems have been built, and Beidou communication technology is also constantly developing into the field of civil technology. Engineering construction in some remote areas is inevitable. In areas without 4G or 5G network coverage, it is very important to ensure convenient communication and safety of construction workers.

[0038] Specifically, images to be transmitted in emergency communications are obtained, and the images include static images and I-frame images in videos. The I-frame images are obtained by processing the video using inter-frame coding technology. The inter-frame coding technology is a well-known technology and will not be described in detail here.

[0039] For example, the image to be transmitted is Figure 2 as shown in .

[0040] Furthermore, the image to be transmitted in the emergency communication is divided into image blocks of the same size, wherein the image block includes a central point and multiple neighborhood points, wherein the central point refers to the pixel point located at the center of the image block, and the remaining pixel points are neighborhood points; the size of the image block is ,in, for An integer in the range.

[0041] For example, when When the image block size is 3×3, Figure 3 As shown in (1) in When the image block size is 5×5, Figure 3 As shown in (2) in When the image block size is 7×7, Figure 3 As shown in (3) in Figure 3 In the image block shown, the gray one is the center point and the white one is the neighborhood point.

[0042] S2. Obtain a key image based on the grayscale values ​​of all center points.

[0043] Specifically, the grayscale values ​​of the center points of all image blocks are combined into a key image; since each size is The image block only includes one center point, so the size of the key image is equal to the size of the image to be transmitted. .

[0044] For example, for Figure 2 The image to be transmitted is shown in FIG. Figure 4 As shown, the size of the key image is equal to the size of the image to be transmitted .

[0045] It should be noted that the acquired key image is equivalent to a downsampled image of the image to be transmitted. Downsampling directly reduces the number of pixels in the image. The reduction in the number of pixels means a reduction in the amount of data that needs to be transmitted, thereby achieving compression of the image to be transmitted, thereby improving transmission efficiency.

[0046] S3. Acquire multiple label values ​​of each neighborhood point in the image block, and obtain multiple supplementary images according to the multiple label values ​​of all neighborhood points.

[0047] It should be noted that although downsampling can achieve compression of the image to be transmitted, it will also lead to the loss of some image features, thereby causing quality loss of the received image; therefore, the image features lost due to downsampling need to be transmitted as supplementary information.

[0048] S301, obtaining a first labeling value of a neighborhood point according to a grayscale value relationship between the neighborhood point and the center point; and obtaining a first supplementary image according to the first labeling values ​​of all neighborhood points.

[0049] Specifically, get the first mark value of the neighborhood point , including: if the gray value of the neighborhood point is greater than the gray value of the center point, the first mark value of the neighborhood point Equal to the first code, otherwise, the first label value of the neighborhood point Equal to the second number; otherwise, it can be based on the first mark value of the neighborhood point , get the gray value relationship between the neighborhood point and the center point: if the first mark value of the neighborhood point is equal to the first code, then the grayscale value of the neighborhood point is greater than the grayscale value of the center point, otherwise, the grayscale value of the neighborhood point is less than or equal to the grayscale value of the center point.

[0050] In one embodiment, the first digital number is 1 and the second digital number is 0; in another embodiment, the first digital number is 0 and the second digital number is 1.

[0051] Exemplarily, when the first digit is 1 and the second digit is 0, for Figure 3 For the image block shown in (1), the grayscale value of the center point is 100. Starting from the neighborhood point in the upper left corner, in a clockwise direction, the grayscale values ​​of each neighborhood point are 103, 98, 100, 101, 104, 91, 106, and 95, respectively. Therefore, the first mark value of each neighborhood point is 1, 0, 0, 1, 1, 0, 1, and 0, respectively.

[0052] Furthermore, the binary number composed of the first mark values ​​of all the neighborhood points in each image block is converted into a decimal number and used as the first representative grayscale value of the image block, and the first representative grayscale values ​​of all the image blocks are composed of a first supplementary image; since each size is The image block has only one first representative gray value, therefore, the first supplementary image has the same size as the key image, and the size of the first supplementary image is equal to the size of the image to be transmitted. .

[0053] For example, for Figure 3 For the image block shown in (1), the first label values ​​of each neighborhood point are 1, 0, 0, 1, 1, 0, 1, 0 respectively. Then the binary number composed of the first label values ​​of all neighborhood points is 10011010. The decimal number after conversion of the binary number is 154. Therefore, the first representative grayscale value of the image block is 154.

[0054] For example, for Figure 2 The image to be transmitted shown in FIG. Figure 5 As shown, the size of the first supplementary image is the same as that of the key image, and the size of the first supplementary image is equal to the size of the image to be transmitted. .

[0055] It should be noted that the first mark value of the neighborhood point can reflect the grayscale value relationship between the neighborhood point and the center point. Therefore, if the first supplementary image is transmitted, the receiver can obtain the grayscale value relationship between the neighborhood point and the center point based on the first supplementary image. Then, when the key image is upsampled to restore the transmitted image, the grayscale value of each neighborhood point after upsampling can be limited according to the grayscale value relationship between the neighborhood point and the center point. Compared with directly restoring the transmitted image based on upsampling the key image, the restored image retains more image features and has better image quality by combining the key image and the grayscale value relationship between the neighborhood point and the center point, which is helpful for subsequent image analysis, target recognition and other operations.

[0056] S302, performing multi-peak Gaussian fitting on the grayscale differences of all neighborhood points in the image to obtain a Gaussian mixture model.

[0057] It should be noted that the first mark value of the neighborhood point can only reflect the grayscale value relationship between the neighborhood point and the center point, and cannot reflect how much greater or less the grayscale value of the neighborhood point is than that of the center point; for example, when the grayscale value of the center point is 100, if it is only known that the grayscale value of the neighborhood point is greater than that of the center point, the grayscale value of the neighborhood point may be 155 grayscale values ​​from 101 to 255; and if the grayscale difference between the neighborhood point and the center point is clear, the grayscale value of the neighborhood point can be determined more accurately; for example, when When the grayscale value of the center point is 100, in addition to knowing that the grayscale value of the neighboring point is larger than that of the center point, it is also known that the grayscale difference between the neighboring point and the center point is in the range of [5,16]. The grayscale value of the neighboring point may be 12 grayscale values ​​from 105 to 116, which is closer to the true grayscale value of the neighboring point than the 155 grayscale values ​​when only knowing that the grayscale value of the neighboring point is larger than that of the center point. Therefore, the present invention obtains a Gaussian mixture model by performing multi-peak Gaussian fitting on the grayscale differences of all neighboring points in the image.

[0058] Specifically, for any neighborhood point, the absolute value of the difference between the grayscale value of the neighborhood point and the grayscale value of the center point is taken as the grayscale difference of the neighborhood point.

[0059] Furthermore, a grayscale difference histogram is constructed based on the grayscale differences of all neighborhood points in the image; the horizontal axis of the grayscale difference histogram is the grayscale differences, and the vertical axis is the frequency of each grayscale difference; the frequency of the grayscale difference is equal to the ratio of the number of all neighborhood points corresponding to the grayscale differences to all neighborhood points in the image.

[0060] Furthermore, the grayscale difference histogram usually contains multiple peaks, each of which corresponds to a type of image feature in the image, such as edge features, texture features, etc. Therefore, a multi-peak Gaussian fitting is performed on the grayscale difference histogram to obtain a Gaussian mixture model (Gaussian Mixture Model, GMM), the Gaussian mixture model It is a probability model whose input is grayscale difference and output is the grayscale difference of the input in Gaussian mixture model. The predicted frequency of It is composed of multiple Gaussian components, each of which has a peak, and there is a trough between two adjacent Gaussian components. Each Gaussian component contains two parameters, namely the mean and the variance, which are used to calculate the predicted frequency of the input grayscale difference in each Gaussian component.

[0061] Exemplarily, the Gaussian mixture model obtained by multi-peak Gaussian fitting like Figure 6 As shown, the figure includes multiple peaks and multiple troughs, and each peak corresponds to a Gaussian component.

[0062] Further, according to Gaussian Components The fitting error and the difference between the probability at the peak and the probability at the trough are used to calculate the Gaussian Components The retention probability of Gaussian Components The retention probability satisfies the expression:

[0063] ;

[0064] In the formula, For the Gaussian Components The probability of retention, For the Gaussian Components The fitting error, Gaussian component The predicted frequency at the peak of It is Gaussian Components With Gaussian Components The predicted frequency of the trough between It is Gaussian Components With Gaussian Components The predicted frequency at the trough between .

[0065] It should be noted that for each Gaussian component, the difference between the predicted frequency at the peak and the predicted frequency at the trough is calculated, which reflects the model's ability to capture the histogram features. The fitting error reflects the accuracy of the model in capturing the histogram features. The retention probability of each Gaussian component is calculated based on the fitting error and probability difference of each Gaussian component. The smaller the fitting error and the greater the predicted frequency difference, the higher the retention probability of the Gaussian component.

[0066] The fitting error of the Gaussian component is realized by comparing the predicted frequency of the Gaussian component with the actual frequency in the grayscale difference histogram. Gaussian Components The fitting error Satisfies the expression:

[0067] ;

[0068] In the formula, For the Gaussian Components The number of all grayscale differences involved, For the Gaussian Components The first The frequency of grayscale differences, For the Gaussian Components The first The grayscale difference is Gaussian Components The predicted frequency of is the absolute value function.

[0069] Further, the one with the highest probability will be retained Gaussian mixture model composed of Gaussian components as the Gaussian mixture model to which the first label value of the neighborhood point belongs .

[0070] in, , is a preset value, and the range of the preset value is ; When the emergency communication has higher requirements for the clarity of the transmitted image, the more Gaussian components are ultimately retained; when the emergency communication has higher requirements for the transmission speed, the fewer Gaussian components are ultimately retained. The specific value can be set according to the actual application scenario and needs. The present invention sets the preset value to 4.

[0071] S303, obtaining each label value of the neighborhood point according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs.

[0072] It should be noted that, when the range of grayscale difference is smaller, the grayscale value of the neighborhood point can be determined more accurately; for example, when the grayscale value of the center point is 100, in addition to knowing that the grayscale value of the neighborhood point is larger than that of the center point, it is also known that the grayscale difference between the neighborhood point and the center point is in the range of [5,16]. The grayscale value of the neighborhood point may be 12 grayscale values ​​from 105 to 116. When the grayscale difference between the neighborhood point and the center point is in the range of [8,11], the grayscale value of the neighborhood point may be 4 grayscale values ​​from 108 to 111, which is closer to the true grayscale value of the neighborhood point than the 12 grayscale values ​​when only the grayscale difference is known to be in the range of [5,16]. Therefore, the present invention obtains each label value of the neighborhood point according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs.

[0073] Specifically, get the first Tagged Value , for Integers in the range of: If the grayscale difference of the neighborhood points belongs to the sub-mixture model on the left, the neighborhood points Tagged Value is equal to the first code, and the sub-mixture model on the left is used as the first The Gaussian mixture model to which the tag value belongs , otherwise, the neighborhood point Tagged Value is equal to the second digit, and the sub-mixing model on the right is taken as .

[0074] Among them, the sub-mixture model is based on the segmentation threshold Gaussian Mixture Model Gaussian mixture model obtained by segmentation is the neighborhood point Gaussian mixture model to which the tag value belongs, segmentation threshold Equal composition The valley between the middle two Gaussian components of all Gaussian components.

[0075] On the contrary, it can be based on the neighborhood point Tagged Value , obtain the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs: if the first Tagged Value If it is equal to the first code, the grayscale difference of the neighborhood points belongs to the sub-mixture model on the left, otherwise, the grayscale difference of the neighborhood points belongs to the sub-mixture model on the right.

[0076] It should be noted that when obtaining the segmentation threshold By combining the components The trough between the two middle Gaussian components of all Gaussian components is used as the segmentation threshold , due to the Gaussian mixture model Depend on Gaussian components, namely Gaussian mixture model It is composed of an even number of Gaussian components, so it is composed of Among all the Gaussian components of , there must be two middle Gaussian components.

[0077] S304: Obtain multiple supplementary images according to multiple label values ​​of all neighborhood points.

[0078] Specifically, all the neighborhood points in each image block are The binary number composed of the tag value is converted into a decimal number and used as the first Represents the gray value, and the Represents the gray value composition Supplementary images; Since each size is The image block has only one represents the gray value, so the The supplementary image is the same size as the key image and The size of the complementary image is equal to the size of the image to be transmitted .

[0079] For example, for Figure 3 For the image block shown in (1), the second label values ​​of each neighborhood point are 1, 1, 1, 1, 1, 0, 1, 1 respectively. Then the binary number composed of the second label values ​​of all neighborhood points is 11111011. The decimal number after conversion of the binary number is 251. Then the second representative grayscale value of the image block is 251.

[0080] For example, for Figure 2 The image to be transmitted shown in FIG. Figure 7 As shown, the size of the second supplementary image is the same as that of the key image, and the size of the second supplementary image is equal to the size of the image to be transmitted. .

[0081] It should be noted that the first The mark value can reflect the range of grayscale difference between the neighborhood point and the center point. The supplementary image is transmitted, and the receiver can The supplementary image obtains the range of grayscale differences between the neighborhood points and the center point, and then when the key image is upsampled to restore the transmitted image, the grayscale value of each neighborhood point after upsampling can be limited according to the range of grayscale differences between the neighborhood points and the center point. Compared with directly restoring the transmitted image by upsampling the key image, the restored image retains more image features and has better image quality by combining the key image and the range of grayscale differences between the neighborhood points and the center point, which is helpful for subsequent image analysis, target recognition and other operations.

[0082] S4. Transmit the key image and multiple supplementary images for emergency communication.

[0083] In step S3, the present invention first obtains the first marking value of the neighborhood point through the grayscale value size relationship between the neighborhood point and the center point, and then obtains the first supplementary image; then, according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs, obtains each marking value of the neighborhood point, and then obtains other supplementary images, thereby continuously narrowing the range of the grayscale difference between the neighborhood point and the center point. The smaller the range of the grayscale difference, the more accurately the grayscale value of the neighborhood point can be determined; therefore, the more supplementary images are received, the more image features retained by the finally restored image, the better the image quality, which is helpful for subsequent image analysis, target recognition and other operations.

[0084] Specifically, since images can be restored based on the key image and different numbers of supplementary images, the image features retained by the final restored image are different, and the image quality is different; therefore, when the key image and the supplementary images are transmitted for emergency communication, the key image and the supplementary images can be obtained while being transmitted; and when the key image and the supplementary images are transmitted for emergency communication, the key image, the first supplementary image, the second supplementary image, ..., the first supplementary image, ... ... Supplementary images, The complementary images are transmitted in order.

[0085] Furthermore, when the situation is urgent, the image can be restored directly based on the currently received key image and multiple supplementary images, or even directly based on the key image, and then image analysis, target recognition and other operations can be performed based on the restored image.

[0086] Among them, when restoring an image only based on the key image, the key image is upsampled and the upsampling result is used as the restored image; when restoring an image based on the key image and the first supplementary image, the key image is upsampled and the upsampling result is adjusted so that in the adjusted upsampling result, the grayscale value size relationship between the neighborhood points and the center point of each image block satisfies the size relationship corresponding to the first mark value of the neighborhood point; when restoring an image based on the key image and multiple supplementary images, the key image is upsampled and the upsampling result is adjusted so that in the adjusted upsampling result, the grayscale value size relationship between the neighborhood points and the center point of each image block satisfies the size relationship corresponding to the first mark value of the neighborhood point, and the grayscale difference between the neighborhood points and the center point of each image block belongs to the Gaussian mixture model corresponding to each mark value of the neighborhood point.

[0087] In addition, the Gaussian mixture model needs to be composed The retention probability is the highest The parameters of the Gaussian components are transmitted as supplementary information.

[0088] It should be noted that the present invention forms a key image with the center points of all image blocks in the image to be transmitted. The obtained key image is equivalent to a downsampled image of the image to be transmitted. The number of pixels in the image is directly reduced through downsampling. The reduction in the number of pixels means that the amount of data to be transmitted is reduced, thereby achieving compression of the image to be transmitted, thereby improving the effectiveness and efficiency of communication transmission.

[0089] It should be further explained that, after acquiring the key image, the present invention first obtains the first mark value of the neighborhood point through the grayscale value relationship between the neighborhood point and the center point, and then obtains the first supplementary image, and then obtains each mark value of the neighborhood point according to the Gaussian mixture model to which the grayscale difference between the neighborhood point and the center point belongs, and then obtains other supplementary images, thereby continuously narrowing the range to which the grayscale difference between the neighborhood point and the center point belongs. After the supplementary image is transmitted, the receiver can obtain the grayscale value relationship between the neighborhood point and the center point and the range to which the grayscale difference between the neighborhood point and the center point belongs according to each supplementary image, and then when the key image is upsampled to restore the transmitted image, the grayscale value of each neighborhood point after upsampling can be limited; compared with directly restoring the transmitted image by upsampling the key image, more image features are retained and the image quality is better, which is helpful for subsequent image analysis, target recognition and other operations.

[0090] An embodiment of the present invention also discloses an emergency communication transmission system based on Beidou satellite, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the emergency communication transmission method based on Beidou satellite according to the present invention is implemented.

[0091] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0092] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0093] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. The emergency communication transmission method based on Beidou satellite is characterized in that: include: The image to be transmitted in emergency communication is divided into image blocks of the same size. The image blocks include a central point and multiple neighborhood points. The central points of all image blocks are combined into a key image. The binary number composed of the tag value is converted into a decimal number as the first Represents the gray value, and the Represents the gray value composition Supplementary images, , is a preset value; performing emergency communication transmission on the key image and at least one supplementary image; Among them, the first mark value of the neighborhood point The method of obtaining is: if the gray value of the neighborhood point is greater than the gray value of the center point, Equal to the first digit, otherwise =Equal to the second digital number; performing multi-peak Gaussian fitting on the grayscale differences of all neighborhood points in the image, and using the obtained Gaussian mixture model as the Gaussian mixture model to which the first label value of the neighborhood point belongs, including: constructing a grayscale difference histogram according to the grayscale differences of all neighborhood points in the image, wherein the horizontal axis of the grayscale difference histogram is the grayscale differences, and the vertical axis is the frequency of the grayscale differences; the frequency of the grayscale difference is equal to the ratio of the number of all neighborhood points corresponding to the grayscale differences to all neighborhood points in the image; performing multi-peak Gaussian fitting on the grayscale difference histogram to obtain a Gaussian mixture model , It is composed of multiple Gaussian components, each of which has a peak, and there is a trough between two adjacent Gaussian components; Gaussian Components The fitting error and the difference between the predicted frequency at the peak and the predicted frequency at the trough are calculated. Gaussian Components The retention probability , , For the Gaussian Components The fitting error, , For the Gaussian Components The number of all grayscale differences involved, For the Gaussian Components The first The frequency of grayscale differences, For the Gaussian Components The first The grayscale difference is Gaussian Components The predicted frequency of is the absolute value function, Gaussian component The predicted frequency at the peak of It is Gaussian Components With Gaussian Components The predicted frequency at the trough between It is Gaussian Components With Gaussian Components The predicted frequency at the trough between Gaussian mixture model composed of Gaussian components as the Gaussian mixture model to which the first label value of the neighborhood point belongs ;in, ; , the neighboring point Tagged Value The method of obtaining is: if the grayscale difference of the neighborhood points belongs to the sub-mixture model on the left, is equal to the first code, and the sub-mixture model on the left is used as the first The Gaussian mixture model to which the tag value belongs ,otherwise is equal to the second digit, and the sub-mixing model on the right is taken as ; The grayscale difference of the neighborhood points refers to the absolute value of the difference between the grayscale values ​​of the neighborhood points and the center point. The sub-mixture model is based on the segmentation threshold Gaussian Mixture Model The segmentation is obtained, is the neighborhood point The Gaussian mixture model to which the tag value belongs, Equal composition The valley between the middle two Gaussian components of all Gaussian components.

2. The Beidou satellite-based emergency communication transmission method according to claim 1, characterized in that: The images to be transmitted in the emergency communication include static images and I-frame images in the video.

3. The Beidou satellite-based emergency communication transmission method according to claim 1, characterized in that: The size of the image block is ,in, for An integer in the range.

4. The Beidou satellite-based emergency communication transmission method according to claim 1, characterized in that: When the key image and at least one supplementary image are transmitted for emergency communication, When , the key image and the first supplementary image are transmitted in the order of When the key image, the first supplementary image, the second supplementary image, ..., the Supplementary images, The complementary images are transmitted in order.

5. The Beidou satellite-based emergency communication transmission method according to claim 1, characterized in that: When the key image and at least one supplementary image are transmitted for emergency communication, each time a key image or a supplementary image is acquired, the acquired key image or supplementary image is transmitted.

6. The Beidou satellite-based emergency communication transmission system is characterized by: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the Beidou satellite-based emergency communication transmission method according to any one of claims 1 to 5 is implemented.

Images