Multi-type meteorological self-recording paper trace identification method based on deep learning

Abstract

The invention relates to the technical field of meteorology, and discloses a multi-type meteorological self-recording paper trace identification method based on deep learning. The deep learning technology is utilized, and a cross-platform computer vision library OpenCV is combined, so that extraction and result quantification of multi-type meteorological self-recording paper trace data are realized; aiming at vertical shaft wind direction / wind speed self-recording paper, adopting a U-net semantic segmentation network to complete trace segmentation extraction; grid extraction is realized by applying a template matching and edge detection algorithm, and then quantification of an extraction result is realized according to related meteorological specifications; the method comprehensively considers the differences of different types of self-recording paper in the aspects of size, color, size, scale, line shape and the like from multiple dimensions of grid extraction, trace extraction, result quantification and the like, highly restores trace information, is suitable for trace digital extraction business processing of wind, rainfall, temperature, air pressure and the like, and has a wide application prospect. The workload and the labor cost can be greatly reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of meteorology, and in particular to a method for recognizing multi-type meteorological self-recording paper traces based on deep learning. Background Art

[0002] Before the advent of automated observation in 2000, my country's historical meteorological observation data were mainly recorded in paper form and manually transcribed. Before the founding of the People's Republic of China, there were various types of self-recording papers, such as air pressure, diurnal temperature, diurnal humidity, French wind speed, vertical axis wind direction, petit dyne wind, R1 / R2 / R3 precipitation, etc. The meteorological data were particularly valuable, but also extremely difficult to apply and process, because they were relatively old, varied, and had different formats and recording methods.

[0003] The trace identification and extraction of historical meteorological self-recording paper can provide timely, accurate and convenient information services for information departments, which is particularly important for carrying out weather monitoring business, enriching meteorological scientific and technological data, and conducting climate forecasts.

[0004] However, the key to trace extraction lies in the accuracy of grid extraction, the accuracy of trace extraction, and the accuracy of the conversion algorithm that converts the image data extracted from the self-recording paper into actual data. Traditional image processing methods mainly focus on relying on single image information such as color information or texture information, which leads to different self-recording papers that cannot effectively remove background noise when the traces become faded, blurred, or smudged, and the extraction results are not accurate.

[0005] The rescue work of precious paper meteorological archive data information is very important. However, in the process of extracting the self-recording paper trace data, due to the differences in types and the characteristics of different self-recording papers, there are the following problems:

[0006] 1. When reconstructing traces using traditional methods, they are easily affected by interference factors such as noise, ink contamination, breakage, pencil scratches, etc., which results in low precision and accuracy of trace extraction;

[0007] 2. The traditional extraction method has its scope of application. When there are many types of self-recording paper and the differences are large, the algorithm that performs well in extracting traces on a certain type of self-recording paper cannot be transferred to other types of self-recording paper;

[0008] 3. Traditional image processing technology can usually only use information such as color and texture of the image, and cannot fully explore the differences in the size, color, size, scale, line shape and other aspects of the self-recording paper, which limits the extraction accuracy;

[0009] 4. In the process of curve tracking, the traditional method cannot accurately track when the steepness of the trace changes greatly, resulting in excessive manual adjustment workload in the later stage. Summary of the invention

[0010] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a multi-type meteorological self-recording paper trace recognition method based on deep learning. The method uses edge gradient features to perform grid template matching for various types of self-recording papers such as air pressure before the founding of the People's Republic of China, diurnal temperature, diurnal humidity, French wind speed, vertical axis wind direction, small dyne wind, R1 / R2 / R3 precipitation, etc., to achieve accurate extraction of self-recording paper grids; the multi-type meteorological self-recording paper trace recognition method based on deep learning, by constructing high-quality data sets of different self-recording papers, combining multiple scale information, and using U-net convolutional neural network to achieve semantic segmentation, eliminate background noise, and complete high-quality and high-precision extraction of traces.

[0011] The present invention provides the following technical solution: a method for identifying multi-type meteorological self-recording paper traces based on deep learning, comprising the following steps:

[0012] (1) Read different types of meteorological scanning image data according to path parameters;

[0013] (2) Parse the read data and determine whether its data format and data type are correct based on the file name;

[0014] (3) Preprocess the read meteorological scanning image data, mainly completing image correction and noise reduction;

[0015] (4) Using a template matching algorithm based on edge gradient features to extract grid traces from the recording paper;

[0016] (5) According to different types of self-recording paper such as temperature, humidity, precipitation, wind direction / speed, etc., a convolutional neural network algorithm based on deep learning is used to mine the spectral and spatial geometric information of the image such as color, texture, shape, etc., to build a training database of different types of self-recording paper and generate a prediction model;

[0017] (6) Using the prediction models generated by different types of self-recording paper, image features are extracted to generate binary images;

[0018] (7) The cross-platform computer vision library OpenCV is used to optimize the trace extraction results, complete noise removal and gap filling, and delete the wrongly extracted traces;

[0019] (8) According to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, and wind speed, the trace extraction results are quantified and a result file is generated.

[0020] Preferably, the specific processing in step 1 includes respectively reading the self-recording paper scanning image data of different stations and different types that require trace extraction.

[0021] Preferably, the specific processing in step 2 includes determining whether the format and data type of the data read in step 1 meet the requirements.

[0022] Preferably, the specific processing in step 3 includes: correcting the image to reduce the data extraction and quantization errors caused by the tilt of the self-recording paper; using a directional filtering-based method to reduce the interference of image noise on the grid extraction and trace extraction of the self-recording paper.

[0023] Preferably, the specific processing in step 4 includes: selecting characteristic characters in the image as a basis for template matching, processing the image characteristic template into discrete points through an edge detection algorithm; achieving template matching by measuring the similarity between the target image and the character template, and completing the extraction of self-recording paper grid traces.

[0024] Preferably, a convolutional neural network algorithm based on deep learning is used to construct a training database of different types of self-recording papers, complete model training, and generate a prediction model.

[0025] Preferably, the step 5 of constructing a training database of different types of self-marking papers refers to manually cutting the image self-marking papers, generating thumbnails, labeling the images, and binarizing the training images by combining color channel attributes with dynamic threshold setting, i.e. generating image training labels.

[0026] The generation of the prediction model in step 5 refers to using the U-net convolutional neural network model, based on the training database generated in step 7, to make full use of the spectral information and spatial geometric information of the image such as color, texture, shape, etc. to generate a training model.

[0027] Preferably, the prediction model generated in step 6 and the different types of self-recording paper images preprocessed in step 3 are used to extract image features and generate a binary image.

[0028] Preferably, the binary feature image generated in step 6 is subjected to a directional filtering algorithm of the computer vision library OpenCV to remove burrs and edge points.

[0029] Preferably, according to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, wind speed, etc., the self-recording paper grid generated in step 5 is used as a calculation basis to quantify the extraction results of each pixel point on the extracted self-recording paper image and generate a result file.

[0030] The present invention has the following beneficial effects:

[0031] 1. This multi-type meteorological self-recording paper trace recognition method based on deep learning, based on the template matching algorithm of edge gradient features, completes the self-recording paper grid extraction. Compared with the traditional image edge detection algorithm, it can adapt to more different types of self-recording papers (different sizes, different styles, and large differences in coordinate systems), and is more universal; the convolutional neural network algorithm based on deep learning adopted by the present invention mines the spectral and spatial geometric information of the image such as color, texture, shape, etc., and eliminates the noise influence of the self-recording paper in the case of faded, blurred, ink smudged, pencil print interference, etc., and highly restores the trace trajectory. Compared with the traditional trace recognition algorithm, the recognition result is more accurate and more universal.

[0032] 2. This multi-type meteorological self-recording paper trace recognition method based on deep learning uses deep learning technology and combines the cross-platform computer vision library OpenCV to realize the extraction and result quantification of multi-type meteorological self-recording paper trace data; the extraction results of this method comprehensively consider various factors such as the size, grid type, scale, trace recording method, color, etc. of different types of self-recording paper. For the temperature, humidity, precipitation and other self-recording papers before the founding of the People's Republic of China, a curve tracking method based on directional filtering is used to complete the trace extraction; for the vertical axis wind direction / wind speed self-recording paper, a U-net semantic segmentation network is used to complete the trace segmentation and extraction; template matching and edge detection algorithms are used to realize grid extraction, and then the extraction results are quantified according to relevant meteorological specifications; this method comprehensively considers the differences in size, color, size, scale, line shape and other aspects of different types of self-recording papers from multiple dimensions such as grid extraction, trace extraction, and result quantification, and highly restores the trace information. It is suitable for the digital extraction and business processing of wind, precipitation, temperature, air pressure and other traces, which can greatly reduce workload and labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the identification method flow of the present invention;

[0034] Figure 2 This is a schematic diagram of wind direction training data of the present invention;

[0035] Figure 3 This is a second schematic diagram of wind direction training data of the present invention;

[0036] Figure 4 This is a schematic diagram of wind speed training data of the present invention;

[0037] Figure 5 This is a second schematic diagram of wind speed training data of the present invention;

[0038] Figure 6 It is a schematic diagram of a binary image of a wind speed trace of the present invention;

[0039] Figure 7It is a schematic diagram of optimizing the binary result of wind speed trace of the present invention;

[0040] Figure 8 This is an example of a trace binary image of the present invention;

[0041] Fig. 9 The filtering template of the present invention;

[0042] Fig.10 This is a schematic diagram of the trace binary filtering result of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] See also Figure 1 , a multi-type meteorological self-recording paper trace recognition method based on deep learning, comprising the following steps:

[0045] (1) Read different types of meteorological scanning image data according to path parameters;

[0046] The specific processing in step 1 includes respectively reading the self-recording paper scanning image data of different stations and different types that need to be subjected to trace extraction;

[0047] The input data contains:

[0048]

[0049]

[0050] (2) parsing the read data and determining whether its data format and data type are correct based on the file name; the specific processing in step 2 includes determining whether the format and data type of the data read in step 1 meet the requirements.

[0051] (3) Preprocessing the read meteorological scanning image data mainly completes image deflection correction and noise reduction; the specific processing in step 3 includes correcting the image to reduce the data extraction and quantization errors caused by the tilt of the self-recording paper; using a method based on directional filtering to reduce the interference of image noise on the grid extraction and trace extraction of the self-recording paper.

[0052] First, a smoothing filter is used to remove noise from the image. The threshold is automatically determined based on local features such as the color information of the image, and the image is binarized using an adaptive binarization method.

[0053] Secondly, the Hough transform algorithm is used to extract straight lines from the table. The horizontal lines of the table are removed using the slope, the slopes of the vertical lines of the table are calculated, and the image is corrected using the average value of the statistical slope.

[0054] The specific steps are as follows:

[0055] 1) Connect the upper endpoints of the grid to obtain the upper line of the grid frame;

[0056] 2) Calculate the inclination angle of the grid line. If the inclination angle is greater than 0.2 degrees, use the affine transformation method to correct the image.

[0057] 3) Affine transformation

[0058] The image should rotate around the center point, but the origin of the image is in the upper left corner. When calculating, the origin of the upper left corner needs to be moved to the center of the image first, and the Y axis needs to be flipped. Let's assume a point (X0, Y0), the image width is W, the height is H, and the point after the origin transformation is (X1, Y1). The transformation is as follows:

[0059]

[0060] The image rotation angle is θ. Suppose the point after the origin is transformed and rotated by the rotation matrix by θ is (X2, Y2). The formula is as follows:

[0061]

[0062] The width of the rotated image is W″ and the height is H″. The formula for transforming from the Cartesian coordinate origin back to the upper left corner is as follows:

[0063]

[0064] In summary, a point (X0, Y0) in the original image can be transformed to the rotated point (X3, Y3) by the following formula:

[0065]

[0066] Similarly, the rotated point (X3, Y3) can be transformed back to the original image (X0, Y0) by the following formula:

[0067]

[0068] (4) Using a template matching algorithm based on edge gradient features to extract grid traces from the recording paper;

[0069] The specific processing in step 4 includes: selecting the characteristic characters in the image as the basis for template matching, processing the image characteristic template into discrete points through the edge detection algorithm; achieving template matching by measuring the similarity between the target image and the character template, and completing the self-recording paper grid trace extraction.

[0070] 1) Make a grid template for your journal

[0071] The production of the grid template is divided into two steps. First, the characteristic characters in the image are selected as the target of template matching. The selected characters can be the time numbers on the edge of the grid, etc.; then, the relative position coordinates of these characteristic characters and the key nodes of the grid are recorded.

[0072] ① Perform a series of rotations, scaling, and pyramid downsampling on the template image to generate a series of templates with different rotation angles and pyramid layers;

[0073] ② Extract the edge of the template and extract the edge points based on the principle of Canny algorithm;

[0074] ③ Calculate the gradient value of the edge point in the x, y direction and the total gradient value;

[0075] ④ Save the x, y gradients corresponding to the edge points, normalize the gradient intensity to eliminate the influence of uneven lighting, and convert the edge point coordinates into relative coordinates relative to the center of gravity.

[0076] 2) Template matching to locate characteristic characters

[0077] ①Determine the similarity measure

[0078] When calculating the edge template matching similarity measure, the edge point gradient vector is used as the matching information. After the template is processed by the edge detection algorithm, the template is converted into a series of points p i =(x i ,y i ) T , i = 1, 2,?, n, corresponding to each point there is a gradient direction vector d i =(t i ,u i ) T .

[0079] The target image is converted into a point (x, y) and its corresponding direction vector e by the same process. x，y =(v x，y , w x，y ) T In the matching process, the transformed template and the sub-image at a certain position of the target image are used to calculate the similarity measure. The gradient vector of the edge point in the template is transformed to d′ i =R(θ)d i, then at a certain position of the target image q = [x, y] T At the sub-image, the similarity metric function is described by the following formula:

[0080]

[0081] In the formula, the average dot product of the gradient vectors of the template and the corresponding edge points of the sub-image is calculated. When the gradient directions of a pair of corresponding edge points are the same, the dot product reaches the maximum value. When the gradient directions are orthogonal, the dot product is zero. The range of the similarity metric value calculated according to the above formula is theoretically the entire real number domain, which is not convenient for judging whether the sub-image meets the requirements, that is, it is difficult to determine a threshold to filter the required results. Therefore, the above formula is normalized:

[0082] (7)

[0084] A desirable feature of the above similarity measure is that when object recognition is based on a threshold S of the similarity measure that a potential match must reach min , there is no need to fully evaluate them. Let S j represents the partial sum of dot products to the jth element of the model. For the matching metric using the normalized dot product sum,

[0085]

[0086] Obviously, all the remaining terms of the sum are ≤ 1. Therefore, if S j <S min -1+j / n, so as long as this condition is met, the summation can stop calculating after the jth element.

[0087] ② Template matching

[0088] Calculate the gradient information of the edge points of the searched image; use two 3*3 matrix operators to convolve with the original image respectively to obtain the gradient values ​​of the horizontal Gx and vertical Gy respectively;

[0089]

[0090] Horizontal and vertical gradient values

[0091] The gradient value of the image is calculated by the following formula:

[0092]

[0093] B) Calculate the gradient information of the edge points of the search image;

[0094] C) pyramid down sampling to obtain edge point gradient information of multi-level pyramid image;

[0095] D) The similarity between the template edge gradient and the target image edge gradient vector is calculated using the Normalized Cross Correlation (NCC) algorithm.

[0096] 3) Extract the mesh

[0097] After determining the characteristic character according to step 2), the grid endpoint is found according to the relative position of the character and the grid line endpoint, and the grid is extracted according to the obtained endpoint.

[0098] (5) According to different types of self-recording paper such as temperature, humidity, precipitation, wind direction / speed, etc., a convolutional neural network algorithm based on deep learning is used to mine the spectral and spatial geometric information of the image such as color, texture, shape, etc., to build a training database of different types of self-recording paper, and generate a prediction model; Using the convolutional neural network algorithm based on deep learning, a training database of different types of self-recording paper is built, model training is completed, and a prediction model is generated;

[0099] The step 5 of constructing a training database of different types of self-marking papers refers to manually completing the cutting of the image self-marking paper, generating small pictures, labeling the images, and using a method combining color channel attributes with dynamic threshold setting to complete the binarization of the training images, that is, generating image training labels;

[0100] Generating a prediction model in step 5 refers to using a U-net convolutional neural network model, based on the training database generated in step 7, to fully utilize the spectral information such as color, texture, shape and spatial geometric information of the image to generate a training model;

[0101] Construct different types of self-recording paper training databases:

[0102] Since the image size of a single self-recording paper is large, considering the hardware facilities and computing efficiency, the self-recording curve image needs to be cropped into smaller images for training. After comprehensive consideration, the image size is set to 320*320, and cross cropping is used for cropping.

[0103] 1) Training image labels:

[0104] Due to the complexity of different types of self-recording paper, manual labeling takes a lot of time. In order to reduce labor costs and save time, the present invention designs a labeling method that combines dynamic thresholds and manual correction. First, the image is dynamically binarized. The principle is as follows: vertical block processing is performed according to the grid lines, and the blue channel values ​​of a single row of pixels are statistically sorted for each image. The value of the middle position is subtracted by 15 as the threshold T of the row. At the same time, other band operations are combined with channel thresholds to binarize a single row of pixels, realize dynamic threshold segmentation of the image, and obtain a preliminary image binary map.

[0105] Binarization formula: (B<T&&(BR)<100)||(R<150)||(BG)>50), where B, G, and R are the pixel values ​​of the blue, green, and red channels of the image pixels.

[0106] Manually correct the preliminary binary image, add the missed extraction part and delete the over-extraction part according to the original image, and obtain the final label after correction. The training image and label correspond one to one, as shown in the figure:

[0107] (6) Using the prediction model generated by different types of self-recording paper, image feature extraction is performed to generate a binary image; using the prediction model generated in step 6 and the different types of self-recording paper images preprocessed in step 3, image feature extraction is performed to generate a binary image

[0108] The binary image extraction method based on Unet network mainly includes the following parts: data preprocessing, U-Net network training, and trace binary image extraction.

[0109] ①Data preprocessing

[0110] The original wind recorder contains two parts: wind direction and wind speed. In order to avoid interference from other factors and obtain wind direction and wind speed information more accurately, the original wind recorder image needs to be cropped into two parts, wind direction and wind speed, and binary images of wind direction and wind speed are extracted respectively.

[0111] Other types of self-recording paper do not need to be cut.

[0112] ②U-net network training

[0113] Select TensorFlow platform as the training environment, build U-net network, and learn and train different types of self-recording paper data. The parameter settings are as follows:

[0114] Lr=0.0001, batch_size=8, epoch=30, save the optimal model.

[0115] ③Wind binary image extraction

[0116] Input the image, call the optimal model of different self-recording paper types respectively, and complete the binary image extraction of different self-recording paper types;

[0117] (7) Using the cross-platform computer vision library OpenCV, the trace extraction results are optimized, noise removal and gap filling are completed, and the wrong traces are deleted; for the binary feature image generated in step 6, the directional filtering algorithm of the computer vision library OpenCV is used to remove burrs and edge points;

[0118] 1) The binary image obtained above is divided into blocks with a width of 100 pixels. The number of ink pixels in each line is counted, and the line with the largest number of ink pixels in the ten lines is taken as the center of the block image. Based on this, the range where the number of ink pixels in two lines is not zero and the distance is greater than 50 is retained, and the pixels outside the range are deleted to eliminate the noise far from the main area of ​​the ink.

[0119] 2) The binary image is extracted by area, and the area with an area less than 200 and a width less than 20 is extracted. The direction of the connected domain is judged by image moment, and the connected domain with an angle α<70 or α>110 is eliminated.

[0120] 3) The missing parts of the trace are supplemented using inherent rules to form an overall ink area, and the upper edge line of the area is extracted to obtain the wind speed trace.

[0121] 4) Horizontal filtering optimization

[0122] The lateral filtering method is used for filtering to remove the false positives on both sides of the trace.

[0123] Filtering method:

[0124] ① Traverse the binary results. If the pixel value of the blue light band of the pixel point is greater than the threshold 150 and is the target pixel of the binary result, set the point as the point to be removed;

[0125] ② Extract the pixel values ​​of the three pixels on the left and right of the point to be removed, and calculate the difference between each pixel value and the point to be removed. If the difference is less than -50, the point is considered to be the edge of the trace and is removed. Otherwise, the point is retained as the trace.

[0126] 5) Remove the mistaken edge points

[0127] There are a few protruding pixels on the edge of the result after horizontal filtering. These protrusions are edge errors. These edges are removed by filtering. The filtering template is shown in the figure:

[0128] (8) According to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, and wind speed, the trace extraction results are quantified, and then a result file is generated. According to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, and wind speed, the self-recording paper grid generated in step 5 is used as the calculation basis to achieve the quantification of the extraction results of each pixel point on the extracted self-recording paper image and generate a result file.

[0129] The technical solution of the present invention is aimed at various types of self-recording papers such as air pressure before the founding of the People's Republic of China, diurnal temperature, diurnal humidity, French wind speed, vertical axis wind direction, small Dyne wind, R1 / R2 / R3 precipitation, etc., and a multi-type meteorological self-recording paper trace recognition method based on deep learning is used. The edge gradient feature is used to perform grid template matching to achieve accurate extraction of self-recording paper grids; by constructing high-quality data sets of different self-recording papers, combining multiple scale information, and using the U-net convolutional neural network, semantic segmentation is achieved, background noise is eliminated, and high-quality and high-precision extraction of traces is completed.

[0130] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0131] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-type meteorological self-recording paper trace recognition method based on deep learning, characterized by: The steps include: (1) Read different types of meteorological scanning image data according to path parameters; (2) Parse the read data and determine whether its data format and data type are correct based on the file name; (3) Preprocess the meteorological scanning image data, mainly completing image correction and noise reduction; (4) Using a template matching algorithm based on edge gradient features to extract grid traces from self-recording paper; (5) According to different types of self-recording paper such as temperature, humidity, precipitation, wind direction / speed, etc., a convolutional neural network algorithm based on deep learning is used to mine the spectral and spatial geometric information of the image such as color, texture, shape, etc., to build a training database of different types of self-recording paper and generate a prediction model; (6) Use the prediction models generated by different types of self-recording paper to extract image features and generate binary images; (7) The cross-platform computer vision library OpenCV is used to optimize the trace extraction results, complete noise removal and gap filling, and delete the wrongly extracted traces; (8) According to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, and wind speed, the trace extraction results are quantified and the result file is generated.

2. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: The specific processing in step 1 includes respectively reading the self-recording paper scanning image data of different stations and different types that need to be subjected to trace extraction.

3. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: The specific processing in step 2 includes determining whether the format and data type of the data read in step 1 meet the requirements.

4. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: The specific processing in step 3 includes: correcting the image to reduce the data extraction and quantization errors caused by the tilt of the self-recording paper; using a method based on directional filtering to reduce the interference of image noise on the grid extraction and trace extraction of the self-recording paper.

5. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: The specific processing in step 4 includes: selecting the characteristic characters in the image as the basis for template matching, processing the image characteristic template into discrete points through the edge detection algorithm; achieving template matching by measuring the similarity between the target image and the character template, and completing the self-recording paper grid trace extraction.

6. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: Using the convolutional neural network algorithm based on deep learning, a training database of different types of self-recording papers was constructed to complete model training and generate a prediction model.

7. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: The step 5 of constructing a training database of different types of self-marking papers refers to manually completing the cutting of the image self-marking paper, generating small pictures, labeling the images, and using a method combining color channel attributes with dynamic threshold setting to complete the binarization of the training images, that is, generating image training labels; The generation of the prediction model in step 5 refers to using the U-net convolutional neural network model, based on the training database generated in step 7, to make full use of the spectral information and spatial geometric information of the image such as color, texture, shape, etc. to generate a training model.

8. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: Using the prediction model generated in step 6 and the different types of self-recording paper images preprocessed in step 3, image features are extracted to generate a binary image.

9. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: For the binary feature image generated in step 6, the directional filtering algorithm of the computer vision library OpenCV is used to remove burrs and edge points.

10. The method for identifying multi-type meteorological self-recording paper traces based on deep learning according to claim 1, characterized in that: According to relevant meteorological specifications such as temperature, humidity, precipitation, wind direction, and wind speed, the self-recording paper grid generated in step 5 is used as the calculation basis to quantify the extraction results of each pixel point on the extracted self-recording paper image and generate a result file.

Images