Meteorological image transmission method and device based on Beidou communication

Abstract

The invention provides a meteorological image transmission method and device based on Beidou communication, and relates to the technical field of artificial intelligence. The method comprises the following steps: acquiring a meteorological detection video containing i meteorological image frames (igt; 1); extracting a specific image descriptor of each meteorological image frame to describe the image frame; generating an overall video descriptor for describing the meteorological detection video according to specific image descriptors of first j (j is a non-zero natural number smaller than i) meteorological image frames in the meteorological detection video; generating a target image descriptor of the ith meteorological image frame by combining the overall video descriptor with a specific image descriptor of the ith meteorological image frame; and obtaining a meteorological observation result of the ith meteorological image frame according to the target image descriptor. According to the embodiment of the invention, the clearness of the meteorological observation result acquisition process can be improved, and the credibility of meteorological observation result acquisition is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for transmitting meteorological images based on Beidou communication. Background Art

[0002] Meteorological monitoring is crucial to many fields such as meteorological research, disaster warning, and agricultural production. Traditional meteorological monitoring mainly relies on meteorological sensors to obtain data such as temperature, humidity, and air pressure, and lacks intuitive image information of meteorological conditions. In recent years, with the development of image acquisition technology, meteorological image monitoring has gradually emerged, but the transmission and analysis of meteorological images face many challenges.

[0003] On the one hand, the amount of meteorological image data is large. When communication resources are limited, such as in remote areas or harsh environments, traditional communication methods are difficult to meet the needs of fast and stable data transmission. On the other hand, existing meteorological image analysis methods have deficiencies in image feature extraction and accuracy of meteorological condition recognition, and cannot efficiently and accurately obtain valuable meteorological observation results from meteorological images. In addition, there is a lack of an effective method that can combine the overall characteristics of meteorological videos and the characteristics of single-frame images for analysis, resulting in the need to improve the comprehensiveness and credibility of meteorological image analysis. Therefore, there is an urgent need for a meteorological image transmission method and device based on Beidou communication to solve the above problems. Summary of the invention

[0004] The object of the present invention is to provide a method and device for transmitting meteorological images based on Beidou communication. The embodiment of the present invention is implemented as follows: In a first aspect, an embodiment of the present invention provides a meteorological image transmission method based on Beidou communication, the method comprising: acquiring a meteorological detection video, the meteorological detection video comprising i meteorological image frames, the number of the meteorological image frames being greater than 1; extracting specific image descriptors that match each of the i meteorological image frames in the meteorological detection video, the specific image descriptors being used to describe the meteorological image frames; obtaining an overall video descriptor based on specific image descriptors that match each of the first j meteorological image frames in the meteorological detection video, the overall video descriptor being used to describe the meteorological detection video, j being a non-zero natural number less than i; obtaining a target image descriptor that matches the i-th meteorological image frame based on the overall video descriptor and the specific image descriptor that matches the i-th meteorological image frame; performing label assignment on the i-th meteorological image frame based on the target image descriptor that matches the i-th meteorological image frame, and determining a meteorological observation result that matches the i-th meteorological image frame; transmitting the meteorological observation result and the meteorological detection video to a receiving end database based on the Beidou short message communication protocol.

[0005] On the other hand, an embodiment of the present invention provides a meteorological image transmission device, including a memory and a processor, wherein the memory stores a computer program that can be executed on the processor, and the processor implements the steps in the above method when executing the program.

[0006] The present invention determines the overall video descriptor matched by the meteorological detection video according to the specific image descriptors matched by the first j meteorological image frames, and then obtains the target image descriptor matched by the ith meteorological image frame according to the overall video descriptor and the specific image descriptor matched by the ith meteorological image frame. The target image descriptor matched by the ith meteorological image frame can reflect not only the characteristics of the ith meteorological image frame, but also the influence of the first j meteorological image frames on the ith meteorological image frame and the global characteristics of the ith meteorological image frame in the meteorological detection video, thereby improving the clarity and accuracy of the target image descriptor, and further enhancing the understandability of the meteorological image frame classification process and the credibility of obtaining meteorological observation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 A schematic diagram of the implementation flow of a meteorological image transmission method based on Beidou communication provided in an embodiment of the present invention.

[0008] Figure 2 It is a hardware entity schematic diagram of a meteorological image transmission device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0009] Figure 1 A schematic diagram of the implementation flow of a method for transmitting meteorological images based on Beidou communication provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method includes: Step 100: Acquire a meteorological detection video, wherein the meteorological detection video includes i meteorological image frames, and the number of the meteorological image frames is greater than one.

[0010] Meteorological detection video is obtained by meteorological image transmission equipment through specific video acquisition equipment. The video can reflect the dynamic changes of meteorological conditions over a period of time. In the meteorological monitoring scene, meteorological conditions will change over time, such as the movement of clouds in the sky, the change of wind direction, the appearance and disappearance of precipitation, etc. These changes will be reflected in the meteorological detection video. Video acquisition equipment is usually installed on platforms such as meteorological monitoring stations, meteorological satellites or drones to ensure that meteorological conditions can be monitored in all directions and from multiple angles; meteorological image frames are static pictures of meteorological detection videos at a certain moment, similar to each frame in a movie. Meteorological image frames contain rich meteorological information, such as the shape, color, thickness of clouds, transparency of the atmosphere, and whether there are meteorological phenomena such as precipitation and lightning. This information is extracted and processed by image analysis technology to provide a basis for meteorological observation and forecasting. i meteorological image frames represent the number of image frames contained in the meteorological detection video, and i is a natural number greater than 1. This shows that the meteorological detection video is not a single image, but consists of multiple continuous image frames, which are arranged in chronological order to form a dynamic video picture. In order to ensure the quality of the collected meteorological detection video, the meteorological image transmission equipment also needs to preprocess the video. Preprocessing includes denoising, enhancement and other operations to remove noise and interference in the video and improve the clarity and contrast of the image frame. The denoising operation uses methods such as mean filtering, median filtering, and Gaussian filtering, which effectively remove salt and pepper noise, Gaussian noise, etc. in the video. The enhancement operation uses methods such as histogram equalization and adaptive histogram equalization, which improve the brightness and contrast of the image frame and make the meteorological information clearer and more visible.

[0011] The formula for mean filtering is: ,in, is the pixel value of the filtered image at (x, y), f(x+i, y+j) is the pixel value of the original image at (x+i, y+j), M N is the size of the filter window. The formula for median filtering is: , Among them, g(x,y) is the pixel value of the filtered image at (x,y), f(x+i,y+j) is the pixel value of the original image at (x+i,y+j), and M N is the size of the filter window, which indicates the median operation. The formula for histogram equalization is: ,in, is the gray level of the equalized image, is the grayscale of the original image, L is the grayscale level of the image, is the original image grayscale The probability density function of .

[0012] Step 200: extracting specific image descriptors that match each of i meteorological image frames in the meteorological detection video, wherein the specific image descriptors are used to describe the meteorological image frames.

[0013] Meteorological image frames are the basic units in meteorological detection videos, similar to each frame in a movie, and contain rich meteorological information, such as the color of the sky, the shape of the clouds, the conditions of precipitation, etc. Meteorological image frames at different times will show different characteristics, which reflect the dynamic changes of meteorological conditions. For example, on a sunny day, the meteorological image frame shows a blue sky and a small amount of white clouds; before a rainstorm, the image frame will show heavy dark clouds and a gloomy sky.

[0014] A specific image descriptor is a digital representation of a meteorological image frame, which converts the complex information in the image frame into a set of representative feature vectors. These feature vectors describe the color distribution, texture features, shape information, etc. of the image frame. For example, the color descriptor uses the mean and standard deviation in the RGB color space to represent the overall color features of the image frame; the texture descriptor extracts the texture information of the image by calculating the gray-level co-occurrence matrix of the image frame; the shape descriptor uses the contour, perimeter, area and other parameters of the object in the image frame to describe the shape features of the object.

[0015] In order to extract specific image descriptors of meteorological image frames, meteorological image transmission equipment uses a variety of technical means. The first is color feature extraction. The meteorological image transmission equipment converts the meteorological image frame into different color spaces, such as RGB, HSV, Lab, etc., and then calculates the statistical features of each channel in the color space, such as mean, standard deviation, histogram, etc. For example, in the RGB color space, the meteorological image transmission equipment calculates the values ​​of the red, green, and blue channels of each pixel, and then counts the mean and standard deviation of these three channels in the entire image frame as the color descriptor.

[0016] Before extracting specific image descriptors, the meteorological image transmission device can preprocess the meteorological image frame to improve the accuracy and stability of feature extraction. Preprocessing operations include denoising, normalization, enhancement, etc. The denoising operation removes noise interference in the image frame, such as salt and pepper noise, Gaussian noise, etc. The meteorological image transmission device uses mean filtering, median filtering, Gaussian filtering and other methods for denoising. The normalization operation adjusts the pixel values ​​of the image frame to a uniform range to eliminate the brightness and contrast differences between different image frames. The meteorological image transmission device uses linear normalization, histogram equalization and other methods for normalization. The enhancement operation improves the clarity and contrast of the image frame and highlights the important features in the image frame. The meteorological image transmission device uses contrast enhancement, sharpening and other methods for enhancement.

[0017] Step 300: obtaining an overall video descriptor according to the specific image descriptors matched by the first j meteorological image frames in the meteorological detection video, wherein the overall video descriptor is used to describe the meteorological detection video, and j is a non-zero natural number less than i; The specific image descriptor is a quantitative representation of the features of a single meteorological image frame, which contains a variety of feature information such as color, texture, shape, etc. For example, for a meteorological image frame containing cumulonimbus clouds, its specific image descriptor will reflect the characteristics of the cloud layer such as darker color, rough texture, and irregular shape. The overall video descriptor is based on these specific image descriptors, comprehensively considering the relationship and changes between the first j image frames, and thus generating a description that can represent the overall characteristics of this video.

[0018] In order to obtain the overall video descriptor, the meteorological image transmission device adopts a variety of technical means. An optional method is based on weighted averaging, which assigns different weights to the specific image descriptors of each image frame according to its importance or the degree of influence on the whole, and then performs weighted summation. This method assumes that each image frame contributes differently to the overall video features, and important image frames should occupy a larger proportion in the overall description.

[0019] For example, suppose the specific image descriptors of the first j meteorological image frames are The corresponding weights are , and satisfies , then the overall video descriptor Calculated by the following formula: In actual applications, weights are determined by a variety of factors. For example, based on the time sequence of image frames, the image frames closer to the current moment have a greater impact on the overall video features, and are therefore given a higher weight. Image frames are also given higher weights based on the content features of the image frames, such as image frames containing important meteorological phenomena (such as heavy rain, typhoons, etc.).

[0020] Another technical approach is based on principal component analysis (PCA). PCA is a data dimension reduction and feature extraction method that converts high-dimensional specific image descriptor data into low-dimensional principal components while retaining the main information of the data. The meteorological image transmission device combines the specific image descriptors of the first j meteorological image frames into a matrix, then performs principal component analysis on the matrix and extracts the principal components as the overall video descriptor.

[0021] Specifically, let the matrix composed of the specific image descriptors of the first j meteorological image frames be X, whose dimension is , where m is the dimension of a specific image descriptor. Through principal component analysis, the covariance matrix of matrix X is obtained , and then solve the eigenvalues ​​and eigenvectors of the covariance matrix C. Select the eigenvectors corresponding to the first p largest eigenvalues ​​to form the projection matrix P, project the matrix X onto the projection matrix P, and get the reduced-dimensional matrix , each column of the matrix Y is a specific image descriptor after dimensionality reduction, and these dimensionality-reduced descriptors are processed comprehensively, such as averaging, to obtain the overall video descriptor.

[0022] In addition to weighted average and principal component analysis methods, meteorological image transmission equipment also uses deep learning methods to obtain overall video descriptors. For example, using models such as recurrent neural networks (RNN) or long short-term memory networks (LSTM), the specific image descriptors of the first j meteorological image frames are sequentially input into the network, and the network automatically learns the sequence relationship and features between image frames, and finally outputs an overall video descriptor. This method can better capture dynamic changes and long-term dependencies in the video.

[0023] When determining the first j meteorological image frames, the meteorological image transmission device needs to be selected according to the actual situation. The value of j is determined according to factors such as the length of the meteorological detection video, the frequency of change, and the accuracy requirements of the analysis. If the video length is short and the changes are fast, j takes a relatively large value to fully consider the characteristics of more image frames; if the video length is long and the changes are relatively slow, j takes a relatively small value to reduce the amount of calculation.

[0024] The overall video descriptor has important application value in meteorological monitoring and analysis. It is used to classify and identify meteorological videos. By comparing the overall video descriptors of different meteorological videos, it is determined whether they belong to the same type of meteorological phenomenon. For example, by comparing the overall video descriptors of videos containing different meteorological scenes such as sunny days, cloudy days, and heavy rain, these videos can be accurately classified. In addition, the overall video descriptor is also used for trend analysis of meteorological changes. By analyzing the changes in the overall video descriptors in different time periods, the development trend of meteorological conditions can be predicted.

[0025] After obtaining the overall video descriptor, the meteorological image transmission device stores the overall video descriptor in the database and associates it with the corresponding meteorological detection video for subsequent query and use. At the same time, in order to ensure the security and credibility of the data, the meteorological image transmission device also needs to back up and encrypt the stored data.

[0026] Step 400: obtaining a target image descriptor matched to the i-th meteorological image frame according to the overall video descriptor and the specific image descriptor matched to the i-th meteorological image frame; In order to obtain the target image descriptor matched by the i-th meteorological image frame, the meteorological image transmission device adopts a variety of technical means. An optional method is feature fusion, which is to directly splice or weightedly combine the overall video descriptor and the specific image descriptor matched by the i-th meteorological image frame. Direct splicing is to connect the two descriptors together in sequence to form a longer feature vector. Weighted combination is to assign different weights to the overall video descriptor and the specific image descriptor matched by the i-th meteorological image frame according to their importance, and then perform weighted summation. If the features of the overall video are more important to the analysis of the current image frame, the weight of the overall video descriptor is appropriately increased; conversely, if the features of the current image frame itself are more critical, the weight of the specific image descriptor matched by the i-th meteorological image frame is increased.

[0027] Another technical approach is based on deep learning, such as using a neural network model to fuse the overall video descriptor and the specific image descriptor that matches the i-th meteorological image frame. A multi-layer perceptron (MLP) model is constructed, and the overall video descriptor and the specific image descriptor that matches the i-th meteorological image frame are used as input. Through model training and learning, the optimal combination of the two descriptors is automatically found, and the target image descriptor is output. During the training process, a large amount of meteorological image frame data and corresponding labels are used, and the model parameters are continuously adjusted through the back propagation algorithm to improve the performance of the model.

[0028] Step 500: performing label assignment on the i-th meteorological image frame according to the target image descriptor matched by the i-th meteorological image frame, and determining the meteorological observation result matched by the i-th meteorological image frame; Meteorological image transmission equipment uses a variety of technical means when assigning labels. One of the optional methods is based on the classifier. A classifier is a model that can classify input feature vectors into different categories. The meteorological image transmission equipment uses a pre-trained classifier and takes the target image descriptor matched by the i-th meteorological image frame as input. The classifier outputs the meteorological category label to which the image frame belongs based on its internal classification rules and trained parameters. Optional classifiers include support vector machine (SVM), decision tree, random forest, etc.

[0029] Taking the support vector machine as an example, the support vector machine finds an optimal hyperplane in the feature space to separate samples of different categories. Let the target image descriptor be x, and its corresponding meteorological category label be y. (Assuming it is a binary classification problem). The goal is to solve the following optimization problem:

[0030] ; Where w is the normal vector of the hyperplane, b is the bias term, is a slack variable used to handle sample misclassification, and C is a penalty parameter used to balance the complexity of the model and the classification error. By solving the above optimization problem, we can obtain the optimal w and b. For the new target image descriptor x, according to to determine the category to which it belongs.

[0031] In practical applications, in order to improve the accuracy of label assignment, meteorological image transmission equipment adopts a multi-classifier fusion method. That is, multiple different types of classifiers are used to classify the target image descriptor of the i-th meteorological image frame, and then a comprehensive judgment is made based on the output results of each classifier. For example, a voting method is used. After each classifier classifies the image frame, the number of votes obtained by each category is counted, and the category with the most votes is the final meteorological observation result.

[0032] Assume that the sky in the ith meteorological image frame acquired by the meteorological image transmission device shows a large gray-white cloud layer, and the target image descriptor reflects that the image color is grayish, the texture is relatively uniform, and the cloud shape is relatively regular. The meteorological image transmission device uses a pre-trained support vector machine classifier and inputs the target image descriptor into the classifier. The classifier determines that the meteorological observation result corresponding to the image frame is "cloudy" based on its trained parameters and rules.

[0033] Step 600: Transmit the meteorological observation results and the meteorological detection video to a receiving end database based on the Beidou short message communication protocol.

[0034] When the meteorological image transmission equipment transmits data based on the Beidou short message communication protocol, it first needs to process the meteorological observation results and meteorological detection videos. Since the bandwidth of Beidou short message communication is limited, the amount of data transmitted at one time cannot be too large, so the meteorological image transmission equipment needs to compress the meteorological detection video to reduce the amount of data. Common video compression methods include H.264, H.265, etc. These methods reduce the amount of video data by removing redundant information in the video, such as time redundancy and spatial redundancy. For example, for a meteorological detection video with a duration of 10 minutes and a resolution of 1080P, the original data volume reaches hundreds of megabytes. After H.264 compression, the data volume will be reduced to tens of megabytes or even smaller.

[0035] For meteorological observation results, meteorological image transmission equipment needs to convert them into a format suitable for Beidou short message communication. Meteorological observation results are usually encoded and represented as a set of numbers or character sequences. For example, "01" represents sunny, "02" represents cloudy, "03" represents rainfall, etc. This reduces the length of data and improves transmission efficiency.

[0036] When transmitting data, the meteorological image transmission equipment needs to follow the provisions of the Beidou short message communication protocol. Beidou short message communication is divided into two modes: RDSS (satellite radio determination service) and RNSS (satellite radio navigation service). The meteorological image transmission equipment needs to select the appropriate mode according to the actual situation. In the RDSS mode, the meteorological image transmission equipment sends short messages to the ground control center through the Beidou satellite, and the ground control center forwards the short messages to the receiving end; in the RNSS mode, the meteorological image transmission equipment communicates directly with the Beidou satellite to realize the sending and receiving of short messages.

[0037] When sending data, the meteorological image transmission equipment needs to divide the meteorological observation results and meteorological detection video data into several short message data packets, and the length of each data packet cannot exceed the maximum length specified by the Beidou short message communication protocol. For example, the maximum length of a single Beidou short message communication is 120 Chinese characters (about 240 bytes), and the meteorological image transmission equipment needs to divide the meteorological observation results and the compressed meteorological detection video data according to this length. At the same time, in order to ensure the integrity and accuracy of the data, the meteorological image transmission equipment needs to add information such as serial number and check code to each data packet. The serial number is used to identify the order of the data packet, and the check code is used to detect whether an error occurs during the transmission of the data packet.

[0038] After receiving the short message data packet sent by the meteorological image transmission device, the receiving end database needs to perform unpacking and verification operations. The receiving end first arranges the data packets in the correct order according to the sequence number, and then uses the check code to verify each data packet. If the verification result shows that the data packet has no errors, it will be combined into a complete meteorological observation result and meteorological detection video data; if the verification result shows that there is an error in the data packet, it is necessary to send a retransmission request to the meteorological image transmission device, requesting the meteorological image transmission device to resend the data packet.

[0039] In order to improve the reliability of data transmission, the meteorological image transmission equipment adopts a retransmission mechanism and a confirmation mechanism. The retransmission mechanism means that after the meteorological image transmission equipment sends a data packet, if it does not receive confirmation information from the receiving end within a certain period of time, it is considered that the data packet transmission has failed and needs to be resent. The confirmation mechanism means that after the receiving end receives a data packet and verifies that it is correct, it sends a confirmation message to the meteorological image transmission equipment, indicating that the data packet has been successfully received.

[0040] In the data transmission link, the Beidou short message communication protocol is followed to ensure the reliability and compatibility of data transmission. The protocol specifies the format of the data packet header, data body, end mark and check code, as follows:

[0041] 1. The data packet header contains information such as protocol type (such as CCTCQ), Beidou receiving device ID, frequency, inbound confirmation application and coding type.

[0042] 2. The data body has different format definitions according to different data types (meteorological observation data, meteorological image data, packet loss retransmission command, etc.). For example, the meteorological observation data body is a calculated meteorological data string, and the meteorological image data body contains information such as time, camera number, number of data packets, and the hexadecimal number of the image data.

[0043] 3. The end mark is fixed to " ", the check code is calculated by checksum method, from the data start mark to the end mark (excluding two characters), the checksum of the remaining characters is accumulated in ASCII code, the accumulated value is encoded in decimal unsigned, the high bit overflows, and the lower two bits are taken.

[0044] Since the amount of data transmitted each time in Beidou short message communication is limited (the communication length range of this transmission protocol is 1-1000 Chinese characters, that is, a maximum of 2000 bytes are transmitted at a time), larger data (such as meteorological image data) needs to be sub-packetized.

[0045] When packetizing, add relevant information about the number of data packets in the data body, such as the total number of data packets and the current number of data packets, so that the receiving end can correctly splice the data. The data packet header contains the protocol type (fixed as CCTCQ), Beidou receiving end device ID (used to identify the device receiving data), frequency (value is 2), inbound confirmation application (set whether confirmation is required according to actual needs), encoding type (such as using compression code to improve transmission efficiency), etc.

[0046] During the transmission process, the data acquisition terminal sends the packaged data to the Beidou satellite, which then forwards the data to the data processing center. If the receiving end (data processing center) finds that the data is lost (judged by the data packet header information and checksum), it sends a lost packet retransmission command to the data acquisition terminal, and the data acquisition terminal retransmits the lost data packet according to the command.

[0047] The data receiving end includes a data receiving and parsing module, a device control instruction generation module, a data storage module and a data analysis module. The data receiving and parsing module receives the data transmitted by the Beidou short message communication module and parses it according to the protocol format. For meteorological observation data, it is directly stored and further analyzed; for meteorological image data, splicing and decompression operations are first performed, the received sub-packet data is spliced ​​in sequence, and then the data is converted back to a binary stream to obtain a directly readable image. The device control instruction generation module generates camera control commands (such as rotation direction and angle), time calibration commands (accurate to seconds), and upload interval and delayed upload commands according to user needs or preset rules. These commands are packaged according to the Beidou short message communication protocol and sent to the data acquisition terminal through the Beidou short message communication module. The data storage module is used to classify and store meteorological observation data and images, and establish a meteorological database for subsequent query, analysis and statistics. The data analysis module uses data mining, machine learning and other algorithms to conduct in-depth analysis of meteorological observation data, such as meteorological change trend prediction and meteorological disaster warning model establishment.

[0048] As an implementation manner, the extracting specific image descriptors that match each of i meteorological image frames in the meteorological detection video includes: Step 210: performing signal purification on the meteorological detection video to generate a purified meteorological detection video; Step 220: extracting mapping features that match each of the i meteorological image frames according to the purified meteorological detection video; Step 230: extracting specific image descriptors that match each of the i meteorological image frames based on the mapping features that match each of the i meteorological image frames.

[0049] In step 210, the meteorological image transmission device performs signal purification on the meteorological detection video to generate a purified meteorological detection video. The purpose of denoising is to remove unnecessary interference information in the video, make the image frame clearer, and facilitate subsequent feature extraction. Optional noise types include salt and pepper noise, Gaussian noise, etc. Salt and pepper noise appears as black and white dots that appear randomly in the image, while Gaussian noise makes the image appear blurred and hazy. In order to remove these noises, the meteorological image transmission device adopts a variety of filtering methods, such as mean filtering, median filtering, and Gaussian filtering. Gaussian filtering is a linear smoothing filtering method based on the Gaussian function. It assigns different weights according to the distance between the pixels in the neighborhood and the center pixel. The closer the pixel is to the center pixel, the greater the weight, and vice versa. The calculation formula of Gaussian filtering is:

[0050] ; in is the standard deviation of the Gaussian function, which controls the smoothness of the filter.

[0051] In step 220, the meteorological image transmission device extracts the mapping features that match each of the i meteorological image frames based on the purified meteorological detection video. The mapping features can reflect the key feature information of the meteorological image frames, which is helpful for subsequent classification, recognition and other operations of the image frames. The meteorological image transmission device extracts mapping features from multiple aspects, such as color features, texture features and shape features. Color features are one of the basic features of an image, which intuitively reflects the overall color distribution of the meteorological image frame. The meteorological image transmission device converts the image frame into different color spaces, such as RGB, HSV, Lab, etc., and then calculates the statistical features of each channel in the color space, such as mean, standard deviation, histogram, etc.

[0052] Texture features reflect the distribution and change of grayscale values ​​in an image. They are used to distinguish different types of meteorological phenomena, such as the texture of clouds and the texture of the ground. Meteorological image transmission equipment uses grayscale co-occurrence matrix, local binary pattern (LBP) and other methods to extract texture information of image frames. Grayscale co-occurrence matrix is ​​a matrix that describes the spatial relationship between grayscale levels in an image. It reflects the texture coarseness, direction and other characteristics of the image. Meteorological image transmission equipment calculates the grayscale co-occurrence matrix at different distances and angles, and then extracts the statistical features of the matrix, such as contrast, correlation, energy and homogeneity, as texture descriptors. Local binary pattern generates a binary pattern by comparing the grayscale values ​​of the central pixel and its neighboring pixels, and then counts the frequency of occurrence of different binary patterns as texture descriptors. Shape features are used to describe the appearance of objects in meteorological image frames, such as the shape of clouds and the shape of rainfall areas. Meteorological image transmission equipment uses edge detection algorithms, such as Canny edge detection algorithm, to extract edge information of objects in image frames, and then calculates parameters such as the contour, perimeter and area of ​​the object as shape descriptors. In addition, meteorological image transmission equipment also uses methods such as moment features and Fourier descriptors to describe the shape characteristics of objects.

[0053] In step 230, the meteorological image transmission device extracts the specific image descriptors that match the i meteorological image frames according to the mapping features that match the i meteorological image frames. The specific image descriptors are more comprehensive and representative descriptions of the meteorological image frames. They integrate various feature information of the image frames and can better reflect the essential features of the image frames. The meteorological image transmission device uses a feature fusion method to integrate mapping features such as color features, texture features, and shape features to generate specific image descriptors. An optional feature fusion method is direct splicing, that is, connecting different types of mapping features in sequence to form a longer feature vector. For example, the color descriptor, texture descriptor, and shape descriptor are connected in sequence to obtain a comprehensive feature vector as a specific image descriptor. Another method is weighted combination, which assigns different weights to different types of mapping features according to their importance, and then performs weighted summation. In the meteorological image classification task, the color feature has a greater impact on the classification, so the weight of the color feature is appropriately increased. In addition, the meteorological image transmission equipment also uses machine learning algorithms, such as principal component analysis (PCA) and linear discriminant analysis (LDA), to reduce the dimension of the mapping features and select features, remove redundant information, and improve the quality and efficiency of specific image descriptors. The meteorological image transmission equipment combines the mapping features of i meteorological image frames into a matrix, and then performs principal component analysis on the matrix to extract the principal components as specific image descriptors. Linear discriminant analysis is a supervised feature extraction method that maximizes the distance between different classes and minimizes the distance within the same class to find the optimal feature projection direction, thereby achieving feature dimension reduction and classification.

[0054] As an implementation, the meteorological image frame includes pixel data of respective matching pixel points; The step of extracting the mapping features matched by each of the i meteorological image frames based on the purified meteorological detection video includes: Step 221: based on the purified meteorological detection video, collecting the target pixel points matched by each of the i meteorological image frames, wherein the target pixel point is the pixel point with the largest pixel data among the multiple pixel points matched by the meteorological image frame; Step 222: for each of the meteorological image frames, taking the target pixel point matched by the meteorological image frame as a reference point, collecting X pixel points along the first diffusion area, and then collecting Y pixel points along the second diffusion area, to generate a pixel point cluster matched by the meteorological image frame; Step 223: extracting mapping features of the meteorological image frame matching based on the pixel data of each pixel point matching in the pixel point cluster of the meteorological image frame matching.

[0055] In step 221, the target pixel represents a position with significant features in the image. The pixel data usually refers to the grayscale value or color value of the pixel. In a grayscale image, the pixel data is the grayscale value of the pixel, which generally ranges from 0 to 255. The larger the value, the brighter the pixel. In a color image, the pixel data is the value of the three channels of RGB. The meteorological image transmission device traverses each pixel in the meteorological image frame, compares the size of their pixel data, and finds the pixel with the largest pixel data as the target pixel. For example, in a purified meteorological image frame, there are some brighter areas in the sky due to light reflection and other reasons. During the traversal process, the meteorological image transmission device will find that the grayscale value of a pixel located in the sky area (assuming it is a grayscale image) is the largest among all the pixels, and then this pixel is determined as the target pixel. In actual operation, the meteorological image transmission device uses a loop statement to traverse the pixel matrix of the image frame row by row and column by column, and records the current maximum pixel data and its corresponding pixel position. Assume that the pixel matrix of the image frame is P(m,n), where m represents the number of rows and n represents the number of columns. During the traversal process, the meteorological image transmission device will compare the values ​​of P(i,j) (i ranges from 0 to m-1, j ranges from 0 to n-1). Assume that the maximum pixel data currently recorded is , the corresponding pixel position is , initially = P(0,0), =(0,0), during the traversal process, if P(i,j)> , then update = P(i,j), =(i,j), after the traversal is completed, It is the location of the target pixel.

[0056] In step 222, the first diffusion area and the second diffusion area are specific areas defined based on the target pixel point, and their shapes and ranges are set according to actual needs. For example, the first diffusion area is a small square area centered on the target pixel point, with a side length of a, and X pixels are collected in this area; the second diffusion area is a larger square area that extends outward from the first diffusion area, with a side length of b (b>a), and Y pixels are collected in this area. When collecting pixel points, the meteorological image transmission device follows a certain order, such as from left to right, from top to bottom. With the target pixel point as the center, For example, when collecting pixels in the first diffusion area, the meteorological image transmission device At the beginning, pixel points are collected in row-first order until X pixel points are collected. When collecting pixel points in the second diffusion area, Y pixel points are collected in a certain order starting from the upper left corner of the area. The collected pixel points form a pixel cluster, which contains the local information around the target pixel point and can reflect the image characteristics of the area.

[0057] In step 223, the mapping feature is a comprehensive representation of the pixel data in the pixel cluster, which can reflect the local features of the meteorological image frame around the target pixel. The meteorological image transmission device uses a variety of methods to extract mapping features, one of which is to calculate the statistical features of the pixel cluster, such as the mean, standard deviation, variance, etc. The mean reflects the average brightness or color value of the pixel cluster, and the standard deviation reflects the degree of discreteness of the pixel data. Assume that the pixel data of the pixel in the pixel cluster are , then the mean of the pixel cluster for , standard deviation for These statistical features are used as part of the mapping features. In addition to the statistical features, the meteorological image transmission device also uses histograms to extract mapping features.

[0058] In practical applications, for example, if there are multiple significant areas in a meteorological image frame, the meteorological image transmission device selects multiple target pixels, collects diffusion areas and extracts mapping features respectively, and then fuses these mapping features to obtain more comprehensive image features. At the same time, the meteorological image transmission device needs to consider the issues of computational efficiency and storage cost to avoid consuming too many resources in the process of extracting mapping features.

[0059] When executing steps 221-223, the meteorological image transmission device also combines image enhancement technology to further improve the quality of the mapping features. For example, before collecting the target pixel points, the purified meteorological image frame is subjected to contrast enhancement processing to make the details in the image clearer, so that the target pixel points can be found more accurately. Contrast enhancement uses methods such as histogram equalization. Histogram equalization adjusts the histogram of the image to make the grayscale distribution of the image more uniform, thereby improving the contrast of the image.

[0060] In addition, when the meteorological image transmission equipment collects pixel clusters and extracts mapping features, the features in the meteorological image frame will change at different scales. The meteorological image transmission equipment uses multi-scale analysis methods, such as scale-invariant feature transform (SIFT) or speeded up robust features (SURF), to extract mapping features at different scales, and then fuses these features to improve the scale invariance of the features. For rotation invariance, the meteorological image transmission equipment uses rotation-invariant feature descriptors, such as local binary patterns (LBP) of circular neighborhoods, which can maintain the stability of features when the image is rotated.

[0061] When executing steps 221-223, the meteorological image transmission device matches the extracted mapping features with known meteorological patterns to further analyze the meteorological conditions. For example, the extracted mapping features are compared with the features of pre-stored meteorological patterns such as cumulonimbus clouds and sunny days. If the matching degree is high, it is determined that the meteorological condition corresponding to the meteorological image frame is the corresponding pattern.

[0062] As an implementation manner, extracting mapping features of the meteorological image frame matching based on pixel data of each pixel point matched in a pixel point cluster of the meteorological image frame matching includes: Step S10: cutting the pixel data matched by the pixel point cluster to generate a plurality of pixel regions, each of which contains pixel data matched by Z pixel points; Step S20: for each of the pixel regions, extracting mapping features of the pixel region matching according to the pixel data of the pixel region matching; Step S30: extracting the matching mapping features of the meteorological image frame according to the matching mapping features of the plurality of pixel regions.

[0063] In step S10, the purpose of cutting pixel data is to divide the pixel cluster into several small, relatively independent regions, so as to analyze the pixel data of each region separately, so as to better capture the local features of the image. The meteorological image transmission device cuts according to certain rules, such as dividing in order of row priority or column priority. Assuming that the pixels in the pixel cluster are arranged in a matrix form, the meteorological image transmission device starts from the upper left corner of the matrix and selects Z pixels to form a pixel area in turn until all pixels are divided into corresponding pixel areas. For example, if the pixel cluster has a total of 100 pixels, set Z = 10, and cut it into 10 pixel areas. This cutting method is similar to dividing a large image into multiple small sub-images, each of which has certain feature information.

[0064] In step S20, the mapping feature is an abstraction and generalization of the pixel data in the pixel area, which can reflect the characteristic information of the area such as color, texture, brightness, etc. The meteorological image transmission device uses a variety of methods to extract the mapping features of the pixel area. An optional method is to calculate the statistical characteristics of the pixel area, such as the mean, standard deviation, variance, etc. The mean reflects the average brightness or color value of the pixel area, and the standard deviation reflects the discrete degree of the pixel data.

[0065] In step S30, the mapping features of each pixel area are integrated to obtain mapping features that can represent the local features of the entire meteorological image frame. The meteorological image transmission device uses a variety of methods for integration, one of which is weighted averaging. The weight is determined based on the importance of the pixel area. For example, a pixel area located in the center area of ​​the image or containing key meteorological features is given a higher weight. Another method is to use principal component analysis (PCA). The meteorological image transmission device combines the mapping features of each pixel area into a matrix, and then performs PCA analysis on the matrix to extract the principal components as mapping features for meteorological image frame matching.

[0066] In practical applications, for example, if there are obvious texture features in the meteorological image frame, the meteorological image transmission device uses a method more suitable for texture analysis to extract the mapping features of the pixel area, such as grayscale co-occurrence matrix, local binary pattern, etc. The grayscale co-occurrence matrix describes the spatial relationship between gray levels in the pixel area. By calculating the statistical characteristics of the grayscale co-occurrence matrix, such as contrast, correlation, energy and homogeneity, the meteorological image transmission device obtains the texture features of the area. The local binary pattern generates a binary pattern by comparing the grayscale values ​​of the central pixel and its neighboring pixels, and then counts the occurrence frequency of different binary patterns as a texture descriptor.

[0067] When cutting pixel data, the meteorological image transmission device also needs to consider the cutting granularity. If the cutting granularity is too small, the number of pixels contained in each pixel area is too small, and sufficient feature information cannot be captured; if the cutting granularity is too large, some local features will be lost. Therefore, the meteorological image transmission device needs to select a suitable Z value based on factors such as the resolution and feature complexity of the meteorological image frame. For example, for high-resolution meteorological image frames, since they contain more detailed information, a smaller Z value is selected to better capture local features; for low-resolution image frames, a larger Z value is selected to reduce the amount of calculation.

[0068] As an implementation manner, extracting the specific image descriptors that match the i meteorological image frames respectively according to the mapping features that match the i meteorological image frames respectively includes: Step 231: for each of the meteorological image frames, calculating the matching influence coefficient of each pixel region in the meteorological image frame; Step 232: integrating the mapping features matched by each pixel region in the meteorological image frame according to the influence coefficient matched by each pixel region in the meteorological image frame to generate a first integrated feature; Step 233: for each pixel region in the meteorological image frame, extract the result of the feature unit alignment multiplication between the feature of the pixel region and the feature of each target pixel region matched by the pixel region; wherein the target pixel region is the next pixel region of the pixel region in the meteorological image frame; Step 234: integrating the bitwise multiplication results of the feature units between the pixel region and each target pixel region matched by the pixel region to generate a second integrated feature of the pixel region match; Step 235: Integrate the second integrated features matched by each pixel region in the meteorological image frame to generate a third integrated feature; Step 236: Integrate the first integrated feature and the third integrated feature to generate a specific image descriptor for matching the meteorological image frame.

[0069] In step 231, the influence coefficient is used to measure the importance of each pixel area in describing the characteristics of the entire meteorological image frame. Different pixel areas contain different meteorological information. For example, in a meteorological image frame containing the sky and the ground, the pixel area of ​​the sky part is more important for judging weather conditions (such as sunny, cloudy), while the pixel area of ​​the ground part also has a certain influence in some cases (such as monitoring the precipitation on the ground). The meteorological image transmission device uses a variety of methods to calculate the influence coefficient. An optional method is based on the variance of the pixel area. The variance reflects the degree of discreteness of the pixel data in the pixel area. The larger the variance, the more drastic the change of the pixel data in the area, and contains more feature information, so a higher influence coefficient is assigned.

[0070] In step 232, the mapping feature is a vector that can reflect the characteristics of the pixel region extracted in the previous step. The purpose of integration is to weight the mapping features of each pixel region according to its influence coefficient to obtain a feature vector that can better represent the entire meteorological image frame. Suppose there are n pixel regions in the meteorological image frame, and the mapping feature of the i-th pixel region is F i , the influence coefficient is , then the first integrated feature Calculated by the following formula: For example, in a weather image frame, there are three pixel regions, and their mapping features are =[1,2,3], =[4,5,6], =[7,8,9], the influence coefficients are = 0.2, , then the first integrated feature: F first =0.2×[1,2,3]+0.3×[4,5,6]+0.5×[7,8,9]=[0.2+1.2+3.5,0.4+1.5+4,0.6+1.8+4.5]=[4.9,5.9,6.9] .

[0071] In step 233, the target pixel region is the next pixel region of the pixel region in the meteorological image frame. Feature unit bitwise multiplication, also known as Hadamard product, refers to multiplying the elements of the corresponding positions of two feature vectors. Suppose the mapping feature of the current pixel region is F current =[ f 1 , f 2 ,⋯, f m ] , the mapping characteristics of the target pixel area are F target =[ g 1 , g 2 ,⋯, g m ] , then the result of multiplying their characteristic units is H=[ f 1 × g 1 , f 2 × g 2 ,⋯, f m × g m ] .

[0072] In step 234, the integration method adopts simple summation or weighted summation. Suppose the result of bitwise multiplication of the feature units of the current pixel area and k target pixel areas is , if a simple summation method is used, the second integrated feature matching of the pixel area For example, the results of the bitwise multiplication of the feature units of the current pixel area and the two target pixel areas are =[1,2,3], H 2 =[4,5,6] , then the pixel area matches the second integrated feature F second-i =[1+4,2+5,3+6]=[5,7,9] .

[0073] In step 235, the integration is also performed by weighted summation.

[0074] In step 236, the integration method is to concatenate or weighted sum the two feature vectors. The weight determination in the weighted summation is adjusted according to the specific application scenario and experimental results. If more attention is paid to the feature information of the pixel region itself, the weight of the first integrated feature is increased; if more attention is paid to the feature association information between adjacent pixel regions, the weight of the third integrated feature is increased.

[0075] In practical applications, for example, when calculating the influence coefficient, in addition to the variance-based method, other factors are also considered, such as the location of the pixel area, the specific meteorological features contained, etc. A higher influence coefficient is assigned to the pixel area located in the center of the image or containing key meteorological features (such as rainstorm clouds, typhoon eye, etc.). When integrating features, more complex integration methods are also used, such as using deep learning models for feature fusion. The first integrated feature and the third integrated feature are input as input to a multi-layer perceptron (MLP) model. Through the training and learning of the model, the optimal feature fusion method is automatically found to generate a more representative specific image descriptor.

[0076] When calculating the bitwise multiplication of feature units and performing feature integration, the meteorological image transmission equipment uses parallel computing methods to improve computing efficiency, using multi-core processors or graphics processing units (GPUs) to accelerate computing. The generated specific image descriptors are compressed to reduce storage space usage.

[0077] In meteorological monitoring and analysis, by comparing the similarity between specific image descriptors of different meteorological image frames, it is determined whether they belong to the same type of meteorological conditions. For example, cosine similarity is used to calculate the similarity between two specific image descriptors. The higher the similarity, the more similar the meteorological conditions of the two image frames are; at the same time, when it is necessary to find other images similar to a specific meteorological image, a quick search is performed based on the specific image descriptor. The meteorological image transmission device stores the specific image descriptors of all meteorological images in the database. When a query image is input, its specific image descriptor is calculated and compared with the descriptors in the database for similarity to find images with higher similarity. In addition, by analyzing the changes in the specific image descriptors of meteorological image frames at different time points, the dynamic change trend of meteorological conditions can be discovered.

[0078] As an implementation manner, the signal purification of the meteorological detection video to generate a purified meteorological detection video includes: Step 211: obtaining a purified coverage range matched by the meteorological detection video, wherein the number of pixel points matched by the purified coverage range is W, where W is an even number; Step 212: for a first pixel point in the weather detection video, take the first pixel point as a starting point, collect pixel data matched by each of the W pixel points in the weather detection video, and obtain the noisy data to be purified matched by the first pixel point; Step 213: determining the robust representative value pixel data in the noisy data to be purified matched by the first pixel point as the purified data matched by the first pixel point; Step 214: Generate the purified meteorological detection video based on the purified data matched to each pixel point in the meteorological detection video.

[0079] In step 211, the purification coverage is a local area used to process the pixels in the meteorological detection video. By operating in this area, the noise is effectively removed. The purpose of selecting a non-even W value is to ensure that the window has a clear center pixel point to facilitate subsequent processing. For example, when W = 3, the purification coverage is a 3×3 window with the center pixel point located in the middle of the window. The meteorological image transmission device selects a suitable W value according to the characteristics of the meteorological detection video and the type of noise. If the noise is concentrated in a local area, the W value is selected to be smaller, such as 3, 5, etc., so that the local noise can be processed more accurately; if the noise distribution is relatively wide, the W value is selected to be larger, such as 7, 9, etc., to cover a larger area for denoising.

[0080] In step 212, this step is to collect pixel data within the determined purification coverage. Taking the 3×3 purification coverage as an example, when the first pixel is located in the upper left corner of the image frame, the meteorological image transmission device will collect pixel data of the pixel and a total of 9 pixels to its right, below and lower right. In actual operation, the meteorological image transmission device needs to ensure that the collected pixels are within the valid range of the image frame. If boundary pixels are encountered, special processing is required, such as mirror extension, zero filling, etc. Mirror extension refers to filling the pixels outside the boundary with pixels at symmetrical positions within the boundary, and zero filling refers to filling the pixels outside the boundary with pixels with a value of 0. Assume that the pixel matrix of the meteorological detection video is P(m,n), where m represents the number of rows, n represents the number of columns, and the coordinates of the first pixel point are (i,j). Then, within the 3×3 purification coverage, the collected pixel data are P(i-1,j-1), P(i-1,j), P(i-1,j+1), P(i,j-1), P(i,j), P(i,j+1), P(i+1,j-1), P(i+1,j), P(i+1,j+1) (assuming that the boundary processing has been completed). These pixel data constitute the noisy data to be purified for the first pixel point matching, which contains noise information and needs to be processed later.

[0081] In step 213, the robust representative value refers to a value in a set of data that can represent the trend of the data set and is insensitive to outliers. The optional robust representative value is the median value. The median value is the value in the middle position after a set of data is arranged in order of size. Taking the 3×3 purification coverage as an example, the meteorological image transmission device arranges the 9 collected pixel data in order from small to large, and then takes the middle value as the purified data matched by the first pixel point. For example, the 9 pixel data collected are [10, 20, 30, 40, 50, 60, 70, 80, 90]. After being arranged in order, the middle value is 50, then 50 is the purified data matched by the first pixel point. The advantage of using the median value as a robust representative value is that it can effectively resist the interference of noise, especially salt and pepper noise. Outliers will have a greater impact on statistics such as the mean, but the impact on the median value is relatively small. Therefore, by taking the median value, the detailed information of the image is retained while removing noise.

[0082] In step 214, the meteorological image transmission device will repeat the operations of step 212 and step 213 for each pixel in the meteorological detection video to obtain the purified data matching each pixel. Then, these purified data are recombined into a new pixel matrix, and this pixel matrix constitutes the purified meteorological detection video. For example, for an m×n meteorological detection video, the meteorological image transmission device will process each pixel in turn to obtain m×n purified data, and arrange these data according to the original position relationship to obtain a purified pixel matrix. Compared with the original video, the noise of the purified meteorological detection video has been effectively removed, and the image frame is clearer, which is conducive to the subsequent feature extraction and analysis of the meteorological image frame.

[0083] In practical applications, for example, when selecting the size of the purification coverage, it is necessary to comprehensively consider the intensity and distribution of the noise and the detailed information of the image. If the noise intensity is large and widely distributed, selecting a larger purification coverage will more effectively remove the noise, but it will cause the detailed information of the image to be lost; if the noise intensity is small and the distribution is concentrated, selecting a smaller purification coverage will better retain the image details while removing the noise. The meteorological image transmission equipment also selects different robust representative value calculation methods according to different types of noise. In addition to the median, other robust statistics are also used, such as the truncated mean. The truncated mean refers to the average value of the remaining data after removing a certain proportion of the maximum and minimum values ​​in a set of data. This method resists the influence of outliers to a certain extent, while taking into account more data information.

[0084] In terms of evaluating the purification effect, the meteorological image transmission device uses some evaluation indicators to measure the quality of the purified meteorological detection video. The optional evaluation indicators are peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM). Peak signal-to-noise ratio is an indicator of the degree of image distortion, which reflects the difference between the original image and the purified image. The calculation formula of PSNR is ,in is the maximum value of the pixels in the image (for 8-bit images, = 255), MSE is the mean square error, calculated as , is the pixel matrix of the original image, is the pixel matrix of the purified image. The higher the PSNR value, the closer the purified image is to the original image, and the better the purification effect. The structural similarity index measures the similarity of images from three aspects: brightness, contrast, and structure. It is more in line with the human eye's perception of image quality. The calculation formula of SSIM is , where x and y are the local areas of the original image and the purified image, respectively. and are the means of x and y respectively, and are the variances of x and y, respectively, is the covariance of x and y, and It is a constant set to avoid the denominator being zero. The closer the SSIM value is to 1, the more similar the structure of the purified image is to the original image, and the better the purification effect is.

[0085] When executing steps 211-214, the meteorological image transmission device also combines other denoising methods for comprehensive processing. For example, it first uses median filtering to perform preliminary denoising on the meteorological detection video, and then uses more complex denoising algorithms, such as wavelet denoising, non-local mean denoising, etc., to further purify the video.

[0086] As an implementation manner, the overall video descriptor is obtained based on the specific image descriptors that match the first j meteorological image frames in the meteorological detection video, including: Step 310: acquiring the influence coefficients of the first j meteorological image frames according to the specific image descriptors that the first j meteorological image frames respectively match; Step 320: Dimension adjustment is performed on the specific image descriptors that match the first j meteorological image frames, so as to generate optimized representation vectors that match the first j meteorological image frames; Step 330: According to the influence coefficients of the first j meteorological image frames, the optimized representation vectors of the first j meteorological image frames are integrated to generate an overall video descriptor.

[0087] In step 310, the influence coefficient is used to measure the importance of each meteorological image frame in the process of constructing the overall video description sub-process. Different meteorological image frames contain different meteorological information, and their contributions to the overall video features are also different. The meteorological image transmission device uses a variety of methods to determine the influence coefficient. One optional method is based on time sequence. The closer the meteorological image frame is to the current moment, the greater the influence on the overall video features, and therefore a higher weight is assigned. For example, assuming that the first j meteorological image frames are arranged in time sequence, the first image frame is the farthest from the current moment, and the jth image frame is the closest to the current moment. The meteorological image transmission device uses a linearly decreasing weight distribution method. Suppose the influence coefficient of the i-th image frame is , For example, when j = 3, the influence coefficient of the first image frame is , the influence coefficient of the second image frame , the influence coefficient of the third image frame Another method is to give higher weights to image frames containing important meteorological phenomena (such as heavy rain, typhoon, etc.) based on the content features of the image frames. The meteorological image transmission equipment identifies the image frames containing key meteorological information by analyzing specific image descriptors and assigns them higher influence coefficients.

[0088] In step 320, the dimension of the specific image descriptor is high and contains a large amount of feature information, but there are some redundant or unimportant features. The purpose of dimensionality adjustment is to reduce the dimension of the feature, remove redundant information, retain the main feature information, and improve the computational efficiency and the expressiveness of the feature. The meteorological image transmission device uses methods such as principal component analysis (PCA) and linear discriminant analysis (LDA) for dimensionality adjustment.

[0089] In step 330, the purpose of integration is to perform weighted combination of the optimized representation vectors of each meteorological image frame according to its influence coefficient, so as to obtain an overall descriptor that can represent the features of the first j image frames of the entire meteorological detection video.

[0090] As an implementation manner, it is characterized in that the step of obtaining a target image descriptor matched by the i-th meteorological image frame based on the overall video descriptor and the specific image descriptor matched by the i-th meteorological image frame comprises: Step 410: Fusing the overall video descriptor and the specific image descriptor matched with the i-th meteorological image frame to obtain a target image descriptor matched with the i-th meteorological image frame.

[0091] The meteorological image transmission equipment adopts a variety of technical means to realize the fusion of the overall video descriptor and the specific image descriptor matched by the ith meteorological image frame. One of the optional methods is weighted fusion. Weighted fusion is to assign different weights to the overall video descriptor and the specific image descriptor matched by the ith meteorological image frame according to their importance, and then perform weighted summation. If the features of the overall video are more important for the analysis of the current image frame, it means that the current meteorological conditions are more related to the meteorological trends of the previous period. For example, when judging the meteorological conditions at a certain moment during continuous rainfall, the previous rainfall trend can provide an important reference. At this time, the weight of the overall video descriptor is appropriately increased; on the contrary, if the features of the current image frame itself are more critical, for example, a special meteorological phenomenon (such as a rare optical phenomenon) suddenly appears at a certain moment, the weight of the specific image descriptor matched by the ith meteorological image frame is increased. Another fusion method is splicing fusion. Splicing fusion is to connect the overall video descriptor and the specific image descriptor matched by the ith meteorological image frame in sequence to form a longer feature vector.

[0092] As another implementation, the meteorological detection video is generated by editing the initial meteorological monitoring video; the method further includes: Step 10: cleaning the first j meteorological image frames in the initial meteorological monitoring video from the initial meteorological monitoring video to generate an optimized initial meteorological monitoring video; In the meteorological monitoring scenario, the first j meteorological image frames of the initial meteorological monitoring video have various problems. For example, when the meteorological monitoring equipment is just started, it takes a certain amount of time to reach a stable working state. During this period, the image frames collected have problems such as uneven brightness and blurred images. In addition, environmental factors such as short-term smoke and dust near the monitoring area also affect the quality of the first j image frames. Suppose that meteorological monitoring is carried out in a mountainous area. When the equipment is started, there is a mountain breeze blowing, which raises some dust. This will cause the sky part in the first few image frames to look gray and cannot truly reflect the meteorological conditions at the time. By cleaning these disturbed first j image frames, the meteorological image transmission device avoids the impact of these inaccurate data on subsequent meteorological analysis.

[0093] In order for the meteorological image transmission device to achieve the cleaning of the first j meteorological image frames from the initial meteorological monitoring video, it is first necessary to determine the value of j. The value of j is determined based on a variety of factors. One method is based on empirical values, and a suitable j value is determined based on previous experience in the same or similar meteorological monitoring scenarios. For example, in multiple meteorological monitoring of a certain area, it was found that the first five image frames after the equipment was started usually had unstable factors, so j was set to 5. Another method is to determine the value of j by performing a preliminary analysis of the initial meteorological monitoring video. The meteorological image transmission device calculates the characteristic indicators of the first several image frames, such as brightness, contrast, image clarity, etc., and compares them with the corresponding indicators of subsequent image frames. If there are obvious differences between the indicators of the first several image frames and the subsequent image frames, and this difference will have an adverse effect on the meteorological analysis, then these image frames are regarded as objects that need to be cleaned. For example, the meteorological image transmission device calculates the average brightness value of each image frame, and assumes that the average brightness value of the kth image frame is , by comparison (n is the total number of image frames in the initial meteorological monitoring video). It is found that the average brightness values ​​of the first j image frames fluctuate greatly, while the average brightness values ​​tend to be stable starting from the j+1th image frame. Then j is determined as the number of image frames that need to be cleaned.

[0094] After determining the value of j, the meteorological image transmission device needs to remove the first j meteorological image frames from the initial meteorological monitoring video. This is achieved by operating the storage structure of the video data. If the initial meteorological monitoring video is stored in the form of a continuous image sequence on a disk or other storage device, the meteorological image transmission device directly skips the first j image files and starts reading and processing data from the j+1th image file. If the video data is stored in a compressed format, such as an optional video encoding format (H.264, H.265, etc.), the meteorological image transmission device needs to decode the video file, locate the position of the first j image frames, and then delete the data of these image frames from the decoded video stream, and finally re-encode the processed video stream to generate an optimized initial meteorological monitoring video.

[0095] In terms of feature extraction of meteorological images, since the first j problematic image frames are removed, the meteorological image transmission device can more accurately extract the features of meteorological image frames, such as color features, texture features, and shape features. For example, when extracting sky color features, if the first j image frames have uneven brightness, the extracted color features will be inaccurate. The optimized video avoids this situation, making the extracted color features more reflective of the true sky color, thus providing a more reliable basis for judging weather conditions (such as sunny, cloudy, overcast, etc.). In terms of weather change trend analysis, the optimized video more clearly shows the dynamic changes in weather conditions.

[0096] Step 20: Obtaining meteorological observation results that match each of the remaining meteorological image frames in the optimized initial meteorological monitoring video; In step 20, the meteorological observation result is a summary description of the meteorological conditions reflected by the meteorological image frames, such as different meteorological types such as sunny, cloudy, rainy, snowy, etc. The optimized initial meteorological monitoring video is the video after removing the first j meteorological image frames, and the remaining meteorological image frames are the legacy meteorological image frames.

[0097] In order to obtain the meteorological observation results matched by each legacy meteorological image frame, the meteorological image transmission device needs to use the meteorological observation algorithm. Specifically, the meteorological image transmission device will use the algorithm to extract the specific image descriptors matched by each legacy meteorological image frame. The specific image descriptors are used to describe the characteristics of the meteorological image frame and contain key information of the meteorological image frame, such as color distribution, texture characteristics, etc. For example, in the meteorological image frame of a sunny day, the color of the sky part is mainly blue, and its specific image descriptor will reflect this color feature. Then, the algorithm will generate an overall video descriptor based on the specific image descriptors matched by the first j legacy meteorological image frames in the optimized initial meteorological monitoring video. The overall video descriptor is used to describe the overall characteristics of the optimized initial meteorological monitoring video. Then, the overall video descriptor and the specific image descriptor matched by the current i-th legacy meteorological image frame are combined to obtain the target image descriptor matched by the i-th legacy meteorological image frame. Finally, according to the target image descriptor matched by the i-th legacy meteorological image frame, label assignment is performed on the i-th legacy meteorological image frame, thereby determining the meteorological observation result matched by the i-th legacy meteorological image frame.

[0098] Step 30: clustering each of the legacy meteorological image frames according to the meteorological observation results matched by each of the legacy meteorological image frames to obtain a plurality of image frame clusters, each of the image frame clusters corresponding to one of the meteorological observation results; In order to cluster the legacy meteorological image frames, the meteorological image transmission equipment needs to use the meteorological observation results that match each of these image frames as the basis for clustering. Meteorological observation results are a key feature of meteorological image frames, which can reflect the essential attributes of the meteorological conditions represented by the image frames. For example, all legacy meteorological image frames marked as "sunny" have similar characteristics in terms of sky color, light intensity, etc.; while legacy meteorological image frames marked as "rainfall" show similarities in features such as cloud morphology and whether there are raindrops in the image. It is based on these similarities that the meteorological image transmission equipment classifies the legacy meteorological image frames to form different image frame clusters.

[0099] In actual operation, meteorological image transmission equipment uses a variety of clustering algorithms to complete this task. Among them, the K-means clustering algorithm is one method. The basic idea of ​​this algorithm is to divide data points into K different clusters in an iterative manner so that the sum of the distances from each data point to the centroid of the cluster to which it belongs is minimized. In the scenario of meteorological image frame clustering, the value of K is the number of types of meteorological observation results. For example, if the meteorological observation results include sunny, cloudy, rainy, and snowy, then K is equal to 4. The meteorological image transmission device first randomly initializes K centroids, each of which represents the center of an image frame cluster. Then, for each legacy meteorological image frame, the meteorological image transmission device calculates its distance to each centroid and assigns it to the image frame cluster represented by the nearest centroid. Next, the meteorological image transmission device updates the centroid of each image frame cluster, that is, calculates the average value of the feature vectors of all legacy meteorological image frames in the cluster. Repeat the above process of assigning and updating the centroid until the centroid no longer changes significantly or reaches the preset number of iterations. Assume that the feature vector of the legacy meteorological image frame is , n is the number of legacy meteorological image frames), the centroid is , then the cluster to which each legacy meteorological image frame x_i belongs is determined by the following formula: ,in Represents the norm of a vector, usually the Euclidean norm is used.

[0100] In addition to the K-means clustering algorithm, the meteorological image transmission device also uses a hierarchical clustering algorithm. Hierarchical clustering algorithms are divided into two types: agglomerative and divisive. The agglomerative hierarchical clustering algorithm starts with each legacy meteorological image frame as a separate cluster, and then gradually merges similar clusters until all legacy meteorological image frames are merged into one cluster or the number of preset clusters is reached. The divisive hierarchical clustering algorithm is the opposite. It starts with all legacy meteorological image frames as a cluster, and then gradually splits the cluster into smaller clusters. In the clustering of meteorological image frames, the meteorological image transmission device selects a suitable hierarchical clustering algorithm according to the characteristics of the meteorological observation results.

[0101] Before clustering, the meteorological image transmission device needs to preprocess the features of the legacy meteorological image frames. Since the features of the meteorological image frames have different scales and ranges, this will affect the performance of the clustering algorithm. Therefore, the meteorological image transmission device normalizes the features and maps the feature values ​​to a uniform range, such as the [0, 1] interval. The normalization methods include minimum-maximum normalization and Z-score normalization. The formula for minimum-maximum normalization is: , where x is the original eigenvalue, and are the minimum and maximum values ​​of the feature, respectively, and x_{norm} is the normalized feature value. The formula for Z-score normalization is: ,in is the mean of the feature, is the standard deviation of the feature.

[0102] The meteorological image transmission equipment also needs to evaluate the clustering results to ensure the quality of clustering. Evaluation indicators include silhouette coefficient, Calinski-Harabasz index, etc. The silhouette coefficient comprehensively considers the closeness of the data point to its cluster and the degree of separation from other clusters. The value range is [-1, 1]. The closer the value is to 1, the better the clustering effect. The Calinski-Harabasz index evaluates the clustering effect by calculating the ratio of the intra-cluster variance to the inter-cluster variance. The larger the value, the better the clustering effect. The meteorological image transmission equipment adjusts and optimizes the clustering results based on these evaluation indicators, such as adjusting the parameters of the clustering algorithm or selecting more appropriate features for clustering.

[0103] Step 40: selecting one of the legacy meteorological image frames from the plurality of image frame clusters for multiple times, and fusing them to obtain a plurality of spliced ​​meteorological images; The meteorological image transmission device performs multiple selection and fusion operations in order to construct a diverse mosaic of meteorological images to cover the characteristic information under different combinations of meteorological observation results, thereby providing a rich data basis for the subsequent evaluation of the clarification analysis approach. For example, in one fusion, the meteorological image transmission device selects a legacy meteorological image frame from the sunny image frame cluster and a legacy meteorological image frame from the rainy image frame cluster, and fuses them into a mosaic of meteorological images containing sunny and rainy features. Through multiple such operations, a mosaic of meteorological images containing various combinations of meteorological observation results is obtained.

[0104] In actual operation, the meteorological image transmission device uses the following technical means to complete this process. First, for selecting legacy meteorological image frames from multiple image frame clusters, the meteorological image transmission device uses a random selection method. It sets an index range for each image frame cluster, and then randomly generates an index within this range, and selects a legacy meteorological image frame from the corresponding image frame cluster based on the index.

[0105] When performing image fusion, the meteorological image transmission device adopts a pixel-based fusion method. This method is to weightedly combine the corresponding pixels of multiple selected legacy meteorological image frames to generate pixel values ​​of the spliced ​​meteorological image. Assume that the meteorological image transmission device selects two legacy meteorological image frames. and , their pixel values ​​at a certain position (x, y) are and , the pixel value P(x,y) of the fused spliced ​​meteorological image at this location is calculated by the following formula: ,in is a weight coefficient ranging from [0,1], which determines the contribution ratio of the two image frames in the fusion process. = 0.5, indicating that the two image frames have the same contribution in the fusion; Near 1 o'clock, The contribution of Near 0 o'clock, The contribution of the meteorological image transmission device is greater. The meteorological image transmission device adjusts the value of the weight coefficient according to actual needs.

[0106] The meteorological image transmission device also uses a feature-based fusion method. This method first extracts the features of the selected legacy meteorological image frames, such as color features, texture features, etc., and then fuses them according to the importance of these features. For example, for color features, the meteorological image transmission device calculates the color histograms of two image frames, and then performs a weighted combination of them to obtain the color histogram of the spliced ​​meteorological image. For texture features, methods such as gray-level co-occurrence matrix are used for extraction and fusion.

[0107] During the multiple selection and fusion processes, the meteorological image transmission equipment needs to record the information of the legacy meteorological image frames used for each fusion to avoid fusing image frames corresponding to completely unrelated meteorological conditions. For example, simply fusing image frames of sunny days and blizzards will cause the spliced ​​meteorological image to lose its practical significance.

[0108] Step 50: for multiple types of clearing analysis approaches, obtaining evaluation values ​​of the clearing analysis approaches under the constraints of each of the stitched meteorological images, wherein the evaluation values ​​are used to evaluate the ratio of the promotion contribution value of the first legacy meteorological image frame in the stitched meteorological image to the promotion contribution value of the stitched meteorological image; In order to obtain the evaluation value of the clarification analysis approach under the constraints of each spliced ​​meteorological image, the meteorological image transmission device needs to process each type of clarification analysis approach and each spliced ​​meteorological image separately. In an embodiment of the present invention, a clarification analysis approach corresponds to an interpretability method. Taking the gradient method as an example, the characteristic changes of the image are analyzed by calculating the gradient value of each pixel in the spliced ​​meteorological image. For a spliced ​​meteorological image, the meteorological image transmission device will use the gradient method to calculate its gradient information, and then calculate the promotion contribution values ​​of the first legacy meteorological image frame and the entire spliced ​​meteorological image under the constraints of the meteorological observation results. Assume that the promotion contribution value of the first legacy meteorological image frame is , the promotion contribution of splicing meteorological images is , then the evaluation value of the gradient method under the constraint of the spliced ​​meteorological image is By formula Calculated.

[0109] Alternatively, the integral gradient method is another clarification analysis approach that calculates the promotion contribution value by integrating the gradient. When using the integral gradient method, the meteorological image transmission device integrates the gradient along the path from the reference image to the spliced ​​meteorological image to obtain the promotion contribution value of each legacy meteorological image frame. The reference image is a blank image or a basic image with specific meteorological characteristics. Similarly, the meteorological image transmission device calculates the promotion contribution value of the first legacy meteorological image frame and the entire spliced ​​meteorological image, and calculates the evaluation value of the integral gradient method under the constraints of the spliced ​​meteorological image according to the above formula.

[0110] Alternatively, the class activation mapping method is used to generate a class activation map to show the areas in the image that have made important contributions to specific meteorological observations. When the meteorological image transmission device uses the class activation mapping method, a heat map reflecting the contribution of each area in the spliced ​​meteorological image to the meteorological observation results is obtained. Then, by analyzing and quantifying the areas corresponding to the first legacy meteorological image frame and the entire spliced ​​meteorological image in the heat map, their contribution values ​​are obtained, and then the evaluation value is calculated.

[0111] Alternatively, the gradient category activation mapping method is used, which combines the ideas of gradient information and category activation mapping. It considers the dual factors of gradient and category activation when calculating the promotion contribution value. When the meteorological image transmission device uses this method, it first calculates the gradient and category activation information of the spliced ​​meteorological image, and then combines these two types of information to calculate the promotion contribution value and evaluation value.

[0112] In actual operation, first, for each clearing analysis path, the meteorological image transmission device needs to pre-process the spliced ​​meteorological image, such as normalization, denoising, etc., to improve the accuracy of the analysis. For example, when using the gradient method, normalizing the image makes the gradient calculation more stable. Then, the meteorological image transmission device will use the corresponding algorithm to implement the clearing analysis path and calculate the promotion contribution value of each legacy meteorological image frame.

[0113] Step 60: obtaining a final evaluation value of the matching of the clarification analysis pathway according to the evaluation value of the clarification analysis pathway under each of the constraints of the spliced ​​meteorological images; The meteorological image transmission device obtains the final evaluation value of the matching of the clearing analysis path in order to comprehensively consider the performance of the analysis path on multiple spliced ​​meteorological images, so as to more accurately evaluate its applicability. Since different spliced ​​meteorological images have different meteorological characteristics and image structures, a single evaluation value cannot fully reflect the performance of the clearing analysis path. Therefore, by calculating the final evaluation value, the meteorological image transmission device objectively compares each clearing analysis path, providing a strong basis for the subsequent determination of the clearing analysis path that is suitable for the initial meteorological monitoring video.

[0114] In actual operation, an optional method used by meteorological image transmission equipment to calculate the final evaluation value is to find the average value. Assume that the evaluation values ​​of a certain clear analysis path under the constraints of n spliced ​​meteorological images are , then the final evaluation value of the matching of the clarification analysis pathway is Calculated by the following formula: For example, if the evaluation values ​​of the gradient method under the constraints of 5 spliced ​​meteorological images are 0.6, 0.7, 0.5, 0.8, and 0.6 respectively, then the final evaluation value of the gradient method is .

[0115] Before calculating the final evaluation value, the meteorological image transmission device uses an outlier detection and processing method, such as a standard deviation-based method, because there are outliers in the evaluation value. The meteorological image transmission device first calculates the mean of the evaluation value. and standard deviation , then the data points whose difference from the mean exceeds k times the standard deviation (k is usually 2 or 3) are considered as outliers and corrected or removed. For example, if the mean of the evaluation value is 0.7, the standard deviation is 0.1, and k is 3, then the points with evaluation values ​​less than 0.4 or greater than 1 can be considered as outliers.

[0116] Step 70: Determine a clarification analysis approach suitable for the initial meteorological monitoring video according to the final evaluation values ​​matched by the various clarification analysis approaches.

[0117] When determining the applicable clearing analysis approach, the meteorological image transmission equipment needs to comprehensively consider the final evaluation values ​​of various clearing analysis approaches. Generally speaking, the higher the final evaluation value, the more accurately the clearing analysis approach can evaluate the ratio of the promotion contribution value of the first legacy meteorological image frame to the promotion contribution value of the spliced ​​meteorological image when analyzing the spliced ​​meteorological image, and the more suitable it is for the initial meteorological monitoring video. For example, if the final evaluation value of the gradient method is 0.8, the final evaluation value of the integral gradient method is 0.7, the final evaluation value of the class activation mapping method is 0.6, and the final evaluation value of the gradient class activation mapping method is 0.75, then from the final evaluation value, the gradient method is the most suitable clearing analysis approach for the initial meteorological monitoring video.

[0118] In actual operation, one method used by meteorological image transmission equipment to determine the clearing analysis path is to directly select the clearing analysis path with the highest final evaluation value. Assume that the final evaluation values ​​of various clearing analysis paths are (m is the number of clarification analysis paths), the meteorological image transmission equipment uses the formula To determine the index of the clarification analysis path with the highest final evaluation value , thus selecting this approach as the applicable analytical method.

[0119] As an implementation manner, the step 50 of obtaining the evaluation value of the clarification analysis approach under the constraints of each of the stitched meteorological images includes: Step 51: for each legacy meteorological image frame in the stitched meteorological image, obtaining a total amount of promotion contribution values ​​between the legacy meteorological image frame and the clarifying analysis pathway, wherein the total amount of promotion contribution values ​​is a summation result of promotion contribution values ​​between each pixel data in the legacy meteorological image frame and the clarifying analysis pathway under the constraint of the meteorological observation result matched by the legacy meteorological image frame; Step 52: summing up the total amount of promotion contribution values ​​between each legacy meteorological image frame in the spliced ​​meteorological image and the clarification analysis path, to obtain a first summing result; Step 53: Divide the total amount of the promotion contribution value between the legacy meteorological image frame and the clearing analysis path by the first addition result, and collect the evaluation value of the clearing analysis path under the constraint of the spliced ​​meteorological image.

[0120] In step 51, the meteorological observation results are obtained by assigning labels to the legacy meteorological image frames through the meteorological observation algorithm, which represents the meteorological conditions reflected by the image frame, such as sunny, cloudy, rainy, etc. Different clarification analysis approaches calculate the promotion contribution value in different ways. Taking the gradient method as an example, this method analyzes the characteristic changes of the image by calculating the gradient value of each pixel in the image. For a legacy meteorological image frame, the meteorological image transmission device will calculate the gradient value of each pixel in the image frame under the constraints of the meteorological observation results, and use these gradient values ​​as the promotion contribution value between the pixel and the gradient method, and then add the promotion contribution values ​​of all pixels to obtain the total promotion contribution value between the legacy meteorological image frame and the gradient method. Assume that a legacy meteorological image frame has N pixels, and the promotion contribution value of the i-th pixel is , then the total contribution value G between the legacy meteorological image frame and the gradient method is expressed by the formula For example, for a 100×100 pixel legacy meteorological image frame, after the meteorological image transmission device calculates the gradient value of each pixel, these 10,000 gradient values ​​are added together to obtain the total amount of promotion contribution value between the image frame and the gradient method.

[0121] In actual operation, the meteorological image transmission device uses the following technical means to calculate the total amount of promotion contribution. First, for the gradient method, the meteorological image transmission device uses the Sobel operator or the Prewitt operator to calculate the gradient value of the pixel. The Sobel operator is an edge detection operator that calculates the gradient components of the pixel in the horizontal and vertical directions by performing a convolution operation on the image, and then combines the two components to obtain the gradient value of the pixel. The Prewitt operator is similar to the Sobel operator and is also used to calculate the gradient of the pixel. For the integral gradient method, the meteorological image transmission device needs to integrate the gradient along the path from the reference image to the legacy meteorological image frame. The reference image is a blank image or a basic image with specific meteorological characteristics. The meteorological image transmission device discretizes this path into multiple steps, calculates the gradient value at each step, and then integrates these gradient values ​​to obtain the promotion contribution value of the pixel. For the category activation mapping method, the meteorological image transmission device generates a heat map that reflects the contribution of each area in the image to the meteorological observation results, and then determines the promotion contribution value of the pixel according to the value of the corresponding pixel in the heat map. For the gradient category activation mapping method, the meteorological image transmission equipment will combine the ideas of gradient information and category activation mapping, first calculate the gradient and category activation information of the image, and then combine these two types of information to determine the promotion contribution value of the pixel point.

[0122] Step 52 is to calculate the total amount of promotion contribution between the entire spliced ​​meteorological image and the clearing analysis path. Assuming that a spliced ​​meteorological image is composed of M legacy meteorological image frames, the total amount of promotion contribution between the jth legacy meteorological image frame and the clearing analysis path is , then the first summation result T is obtained by the formula For example, a spliced ​​meteorological image consists of three legacy meteorological image frames, and the total contribution values ​​between them and the gradient method are 100, 150, and 200, respectively. Then the first sum is 100+150+200 = 450.

[0123] In step 53, assuming that the total amount of promotion contribution between the legacy meteorological image frame and the clearing analysis path is G1, and the first sum result is T, then the evaluation value E of the clearing analysis path under the splicing meteorological image constraint is calculated by the formula For example, the total contribution value between the legacy meteorological image frame and the gradient method is 100, and the first sum is 450. Then the evaluation value of the gradient method under the constraint of the spliced ​​meteorological image is .

[0124] The evaluation value reflects the relative importance of the legacy meteorological image frame in the spliced ​​meteorological image. The higher the evaluation value, the greater the contribution of the legacy meteorological image frame to the meteorological observation results represented by the spliced ​​meteorological image; the lower the evaluation value, the smaller the contribution of the legacy meteorological image frame.

[0125] As another implementation, the method is performed by a meteorological observation algorithm, and the method further includes: Step a: Obtain an algorithm tuning template, wherein the algorithm tuning template includes i meteorological image frames, each of the meteorological image frames is marked with a meteorological observation label, and the number of the meteorological image frames is greater than 1.

[0126] The algorithm training template is the training data used to train the meteorological observation algorithm. There are many ways for the meteorological image transmission device to obtain the algorithm training template. One optional way is to collect meteorological image frames from the meteorological monitoring database. The meteorological monitoring database usually stores a large amount of meteorological image data, which is collected by various meteorological monitoring equipment, such as meteorological satellites, meteorological radars, ground cameras, etc. The meteorological image transmission device selects i meteorological image frames from the database according to certain rules, such as selecting image frames in different time periods and different geographical locations to ensure that the algorithm training template is diverse and representative. During the screening process, the meteorological image transmission device also needs to add corresponding meteorological observation labels to each meteorological image frame. These labels are manually marked according to the image content, and are also automatically generated by existing reliable meteorological observation systems.

[0127] Step b: extracting specific image descriptors that match each of the i meteorological image frames in the algorithm adjustment template through the meteorological observation algorithm, wherein the specific image descriptors are used to describe the meteorological image frames; Step c: Generate an overall video descriptor by using the meteorological observation algorithm according to the specific image descriptors that match the first j meteorological image frames in the algorithm training template, wherein the overall video descriptor is used to describe the algorithm training template, and j is a non-zero natural number less than i; Step d: generating a target image descriptor for matching the i-th meteorological image frame according to the overall video descriptor and the specific image descriptor for matching the i-th meteorological image frame; Step e: performing label assignment on the i-th meteorological image frame according to the target image descriptor matched by the i-th meteorological image frame through the meteorological observation algorithm to obtain the predicted meteorological observation result matched by the i-th meteorological image frame; Step f: Debug the meteorological observation algorithm based on the predicted meteorological observation results and meteorological observation labels matched by the i-th meteorological image frame to generate a precisely calibrated meteorological observation algorithm, which is used to perform label assignment on meteorological image frames in meteorological detection videos.

[0128] The principles of steps b~e can refer to the aforementioned steps 200~600 and will not be repeated here. When debugging the algorithm, the meteorological image transmission device calculates the error between the predicted meteorological observation results and the meteorological observation labels. An optional error calculation method is to use a loss function. For example, a cross entropy loss function can be used. Assuming that there are C categories of meteorological observation results, the true label of the i-th meteorological image frame is ( is a C-dimensional one-hot encoding vector, where only the element corresponding to the true category is 1 and the rest are 0). The prediction result is ( is a C-dimensional vector, each element represents the probability that the image frame belongs to the corresponding category), then the cross entropy loss function L is calculated by the formula Calculated, where and They are and For example, if the true label of a weather image frame is = [1, 0, 0] (indicates sunny day), prediction result = [0.8, 0.1, 0.1], then the cross entropy loss .

[0129] After obtaining the error, the meteorological image transmission device needs to adjust the parameters of the meteorological observation algorithm according to the error. If the meteorological observation algorithm is based on a neural network, the parameter adjustment method is the back propagation algorithm. The back propagation algorithm calculates the gradient of the loss function for each parameter in the neural network, and then updates the parameter according to the direction and size of the gradient. Suppose a parameter in the neural network is , the loss function is L, then the parameter update formula is ,in is the learning rate, which controls the step size of parameter updates. For example, if , , learning rate = 0.01, then the updated parameters .

[0130] Figure 2 A hardware entity diagram of a meteorological image transmission device provided by an embodiment of the present invention, such as Figure 2 As shown, the hardware entity of the meteorological image transmission device A includes: a processor A1 and a memory A2, wherein the memory A2 stores a computer program that can be run on the processor A1, and the processor A1 implements the steps in the method of any of the above embodiments when executing the program.

[0131] The above description is only an implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for transmitting meteorological images based on Beidou communication, characterized in that: The method comprises: Collecting a meteorological detection video, wherein the meteorological detection video includes i meteorological image frames, and the number of the meteorological image frames is greater than 1; Extracting specific image descriptors that match each of i meteorological image frames in the meteorological detection video, wherein the specific image descriptors are used to describe the meteorological image frames; According to the specific image descriptors matched by the first j meteorological image frames in the meteorological detection video, an overall video descriptor is obtained, wherein the overall video descriptor is used to describe the meteorological detection video, and j is a non-zero natural number less than i; Obtaining a target image descriptor matched to the i-th meteorological image frame according to the overall video descriptor and the specific image descriptor matched to the i-th meteorological image frame; According to the target image descriptor matched by the i-th meteorological image frame, label assignment is performed on the i-th meteorological image frame to determine the meteorological observation result matched by the i-th meteorological image frame; The meteorological observation results and the meteorological detection video are transmitted to a receiving end database based on the Beidou short message communication protocol.

2. The method for transmitting meteorological images based on Beidou communication as claimed in claim 1, characterized in that: The extracting of specific image descriptors that match i meteorological image frames in the meteorological detection video includes: Performing signal purification on the meteorological detection video to generate a purified meteorological detection video; Extracting mapping features matching each of the i meteorological image frames according to the purified meteorological detection video; According to the mapping features that respectively match the i meteorological image frames, specific image descriptors that respectively match the i meteorological image frames are extracted.

3. The method for transmitting meteorological images based on Beidou communication as claimed in claim 2, characterized in that: The meteorological image frame includes pixel data of respective matching pixel points; The step of extracting the mapping features matched by each of the i meteorological image frames based on the purified meteorological detection video includes: According to the purified meteorological detection video, target pixel points that are matched by each of the i meteorological image frames are collected, and the target pixel point is the pixel point with the largest pixel data among the multiple pixel points matched by the meteorological image frame; For each of the meteorological image frames, taking the target pixel point matched by the meteorological image frame as a reference point, collecting X pixel points along the first diffusion area, and then collecting Y pixel points along the second diffusion area, to generate a pixel point cluster matched by the meteorological image frame; The mapping features of the meteorological image frame matching are extracted based on the pixel data of each pixel point matched in the pixel point cluster of the meteorological image frame matching.

4. The method for transmitting meteorological images based on Beidou communication as claimed in claim 3, characterized in that: The step of extracting the mapping features of the meteorological image frame matching based on the pixel data of each pixel point matched in the pixel point cluster of the meteorological image frame matching comprises: Cutting the pixel data matched by the pixel point cluster to generate a plurality of pixel regions, each of which contains pixel data matched by Z pixel points; For each of the pixel regions, extracting mapping features of the pixel region matches according to the pixel data matched by the pixel region; The mapping features matched by the meteorological image frame are extracted based on the mapping features matched by the plurality of pixel regions.

5. The method for transmitting meteorological images based on Beidou communication as claimed in claim 4, characterized in that: The extracting, based on the mapping features that the i meteorological image frames respectively match, the specific image descriptors that the i meteorological image frames respectively match, comprises: For each of the meteorological image frames, calculating the influence coefficient of each pixel region in the meteorological image frame; Integrate the mapping features matched by each pixel region in the meteorological image frame according to the influence coefficient matched by each pixel region in the meteorological image frame to generate a first integrated feature; For each pixel region in the meteorological image frame, extract the result of the feature unit bitwise multiplication between the feature of the pixel region and the feature of each target pixel region matched by the pixel region; wherein the target pixel region is the next pixel region of the pixel region in the meteorological image frame; Integrate the bitwise multiplication results of the feature units between the pixel region and each target pixel region matched by the pixel region to generate a second integrated feature of the pixel region match; Integrate the second integrated features matched by each pixel region in the meteorological image frame to generate a third integrated feature; The first integrated feature and the third integrated feature are integrated to generate a specific image descriptor for matching the meteorological image frame.

6. The method for transmitting meteorological images based on Beidou communication as claimed in claim 2, characterized in that: The step of performing signal purification on the meteorological detection video to generate a purified meteorological detection video includes: Obtaining a purified coverage range matched by the meteorological detection video, wherein the number of pixel points matched by the purified coverage range is W, where W is an even number; For a first pixel point in the meteorological detection video, taking the first pixel point as a starting point, collecting pixel data matched by each of the W pixel points in the meteorological detection video, and obtaining the noisy data to be purified matched by the first pixel point; Determine the robust representative value pixel data in the noisy data to be purified matched by the first pixel point as the purified data matched by the first pixel point; The purified meteorological detection video is generated based on the purified data matched to each pixel point in the meteorological detection video.

7. The method for transmitting meteorological images based on Beidou communication according to claim 1, characterized in that: The method of obtaining the overall video descriptor based on the specific image descriptors that are matched to the first j meteorological image frames in the meteorological detection video comprises: Obtaining influence coefficients of the matches of the first j meteorological image frames according to the specific image descriptors that the first j meteorological image frames respectively match; Performing dimension adjustment on the specific image descriptors that match the first j meteorological image frames, respectively, to generate optimized representation vectors that match the first j meteorological image frames, respectively; According to the influence coefficients of the respective matches of the first j meteorological image frames, the optimized representation vectors of the respective matches of the first j meteorological image frames are integrated to generate the overall video descriptor; The step of obtaining a target image descriptor matched by the i-th meteorological image frame based on the overall video descriptor and the specific image descriptor matched by the i-th meteorological image frame comprises: The overall video descriptor and the specific image descriptor matched with the i-th meteorological image frame are fused to obtain a target image descriptor matched with the i-th meteorological image frame.

8. The method for transmitting meteorological images based on Beidou communication according to claim 1, characterized in that: The meteorological detection video is generated by editing the initial meteorological monitoring video; the method further comprises: Cleaning the first j meteorological image frames in the initial meteorological monitoring video from the initial meteorological monitoring video to generate an optimized initial meteorological monitoring video; Obtaining meteorological observation results that match each of the remaining meteorological image frames in the optimized initial meteorological monitoring video; Clustering each of the legacy meteorological image frames according to the meteorological observation results matched by each of the legacy meteorological image frames to obtain a plurality of image frame clusters, each of the image frame clusters corresponding to one of the meteorological observation results; Selecting one of the legacy meteorological image frames from the plurality of image frame clusters for multiple times and fusing them to obtain a plurality of spliced ​​meteorological images; For multiple types of clearing analysis approaches, obtaining evaluation values ​​of the clearing analysis approaches under the constraints of each of the stitched meteorological images, wherein the evaluation values ​​are used to evaluate the ratio of the promotion contribution value of the first legacy meteorological image frame in the stitched meteorological image to the promotion contribution value of the stitched meteorological image; According to the evaluation value of the clarification analysis pathway under each of the stitched meteorological image constraints, obtaining a final evaluation value of the clarification analysis pathway matching; Determining a clarification analysis approach suitable for the initial meteorological monitoring video according to the final evaluation values ​​matched by the various clarification analysis approaches; The step of obtaining the evaluation value of the clarification analysis approach under the constraints of each of the stitched meteorological images comprises: For each legacy meteorological image frame in the stitched meteorological image, a total amount of promotion contribution values ​​between the legacy meteorological image frame and the clarifying analysis path is obtained, wherein the total amount of promotion contribution values ​​is a summation result of promotion contribution values ​​between each pixel data in the legacy meteorological image frame and the clarifying analysis path under the constraint of the meteorological observation result matched by the legacy meteorological image frame; Adding the total amount of promotion contribution values ​​between each legacy meteorological image frame in the spliced ​​meteorological image and the clarification analysis path to obtain a first summation result; The total amount of the promotion contribution value between the legacy meteorological image frame and the clearing analysis approach is divided by the first summation result, and the evaluation value of the clearing analysis approach under the constraint of the spliced ​​meteorological image is collected.

9. The method for transmitting meteorological images based on Beidou communication according to claim 1, characterized in that: The method is performed by a meteorological observation algorithm, and the method further comprises: Obtain an algorithm tuning template, wherein the algorithm tuning template includes i meteorological image frames, each of the meteorological image frames is marked with a meteorological observation label, and the number of the meteorological image frames is greater than 1; Extracting specific image descriptors that match each of i meteorological image frames in the algorithm adjustment template through the meteorological observation algorithm, wherein the specific image descriptors are used to describe the meteorological image frames; Generate an overall video descriptor by using the meteorological observation algorithm according to the specific image descriptors that match the first j meteorological image frames in the algorithm training template, wherein the overall video descriptor is used to describe the algorithm training template, and j is a non-zero natural number less than i; Generate a target image descriptor for matching the i-th meteorological image frame according to the overall video descriptor and the specific image descriptor for matching the i-th meteorological image frame; Performing label assignment on the i-th meteorological image frame according to the target image descriptor matched by the i-th meteorological image frame through the meteorological observation algorithm to obtain a predicted meteorological observation result matched by the i-th meteorological image frame; The meteorological observation algorithm is debugged based on the predicted meteorological observation results and meteorological observation labels matched by the i-th meteorological image frame to generate a precisely calibrated meteorological observation algorithm, which is used to perform label assignment on meteorological image frames in meteorological detection videos.

10. A meteorological image transmission device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the program, the steps in the method according to any one of claims 1 to 9 are implemented.

Images