Information processing method, device and computer readable storage medium

Abstract

The embodiment of the present application discloses an information processing method, which includes: obtaining a sample image, and training the sample image to generate a text detection model; wherein the sample image is a picture with text information; obtaining sample information, and training the sample information to generate an information recognition model; obtaining a picture to be recognized associated with identification information; recognizing the picture to be recognized based on the text detection model to obtain a text information set; processing the text information set to determine a target vector; and obtaining target text information associated with the identification information based on the information recognition model and the target vector. The embodiment of the present application also discloses an information processing device and a computer-readable storage medium. The embodiment of the present application realizes the rapid extraction of text public opinion information in image public opinion information.

Description

Technical Field

[0001] The present application relates to the field of electronics and information technology, and in particular to an information processing method, device and computer-readable storage medium. Background Art

[0002] With the development of information technology, it has become a mainstream in society for users to search for online public opinion information carried by pictures on the entire network based on the key information they input. Among the related technologies, image recognition technology is generally used, which has a cumbersome processing process and poor real-time performance. It is used to identify online public opinion information carried by pictures and obtain picture public opinion information, which takes a long time to obtain picture public opinion information. Summary of the invention

[0003] In order to solve the above technical problems, the embodiments of the present application hope to provide an information processing method, device and computer-readable storage medium, which can realize the rapid extraction of text public opinion information in image public opinion information.

[0004] To achieve the above purpose, the technical solution of this application is implemented as follows:

[0005] An information processing method, the method comprising:

[0006] Obtaining a sample image, and training the sample image to generate a text detection model; wherein the sample image is an image with text information;

[0007] Acquire sample information, and train the sample information to generate an information recognition model;

[0008] Obtain the image to be identified that is associated with the identification information;

[0009] Identify the image to be identified based on the text detection model to obtain a text information set;

[0010] Processing the text information set to determine a target vector;

[0011] Based on the information recognition model and the target vector, target text information associated with the identification information is obtained.

[0012] Optionally, obtaining the to-be-recognized picture associated with the identification information includes:

[0013] Obtaining a preprocessed image associated with the identification information;

[0014] Perform image enhancement on the preprocessed image to obtain the image to be identified.

[0015] Optionally, obtaining a preprocessed image associated with the identification information includes:

[0016] Acquire an image to be identified that is associated with the identification information; wherein the image to be identified is an image having multiple frames of static pictures;

[0017] The image to be identified is analyzed to obtain the pre-processed image.

[0018] Optionally, the parsing process of the image to be identified to obtain the pre-processed image includes:

[0019] Decomposing the image to be recognized to obtain an intermediate picture set; wherein the intermediate picture set includes at least two intermediate pictures;

[0020] Similarity calculation is performed on the at least two intermediate pictures, and intermediate pictures are selected from the intermediate picture set to obtain the pre-processed picture.

[0021] Optionally, performing image enhancement on the preprocessed image to obtain the image to be identified includes:

[0022] Acquire a sample image, and train the sample image to generate a direction detection model;

[0023] Performing data enhancement on the preprocessed image using a grayscale transformation algorithm to generate an enhanced image;

[0024] The direction type of the enhanced image is determined based on the direction detection model, and the enhanced image is rotated based on the direction type and a preset rotation strategy to obtain the image to be identified.

[0025] Optionally, the identifying the image to be identified based on the text detection model to obtain a text information set includes:

[0026] Detect the image to be identified based on the text detection model to obtain text line information in the image to be identified;

[0027] The text line information is classified to obtain the text information set.

[0028] Optionally, the processing the text information set to determine the target vector includes:

[0029] Obtain the length value and height value corresponding to each text line information in the text information set;

[0030] Based on the length value and the height value corresponding to each text line information, obtaining the ratio value corresponding to each text line information;

[0031] Based on the ratio value corresponding to each text line information, obtaining a target ratio value;

[0032] A vector is constructed based on the target ratio value to obtain the target vector.

[0033] Optionally, obtaining target text information associated with the identification information based on the information recognition model and the target vector includes:

[0034] The target vector is forward calculated based on the information recognition model to obtain the target text information.

[0035] Optionally, after obtaining the target text information associated with the identification information based on the information recognition model and the target vector, the method further includes:

[0036] Build a sensitive corpus;

[0037] The target text information is filtered based on the sensitive corpus to obtain target security information.

[0038] An information processing device, comprising: a processor, a memory and a communication bus;

[0039] The communication bus is used to realize the communication connection between the processor and the memory;

[0040] The processor is used to execute the program in the memory to implement any of the above-mentioned information processing methods.

[0041] A computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the above-mentioned information processing method.

[0042] The information processing method, device and computer-readable storage medium provided in the embodiments of the present application obtain sample images, and train the sample images to generate a text detection model; wherein the sample images are images with text information; obtain sample information, and train the sample information to generate an information recognition model; obtain images to be recognized that are associated with identification information; recognize the images to be recognized based on the text detection model to obtain a text information set; process the text information set to determine a target vector; obtain target text information associated with the identification information based on the information recognition model and the target vector; in this way, the images to be recognized that are associated with the identification information are processed by the trained and generated text detection model to obtain a text information set, and the text information set is converted into a target vector, and the target vector is processed by the trained and generated information recognition model to obtain the target text information associated with the identification information, instead of using image recognition technology with cumbersome processing procedures and poor real-time performance to process network public opinion information with images as carriers as in the related art, thereby realizing the rapid extraction of text public opinion information in image public opinion information. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A flowchart of an information processing method provided in an embodiment of the present application;

[0044] Figure 2 A flowchart of another information processing method provided in an embodiment of the present application;

[0045] Figure 3 A flowchart of another information processing method provided in an embodiment of the present application;

[0046] Figure 4 A schematic diagram of a picture to be identified containing text line information provided by an embodiment of the present application;

[0047] Figure 5 A schematic diagram of the structure of a full-scale public opinion technology system processing provided in an embodiment of the present application;

[0048] Figure 6 A schematic diagram of the structure of an information processing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0050] It should be understood that the "embodiments of the present application" or "the aforementioned embodiments" mentioned throughout the specification mean that specific features, structures or characteristics related to the embodiments are included in at least one embodiment of the present application. Therefore, "in the embodiments of the present application" or "in the aforementioned embodiments" appearing throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. In the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The above-mentioned serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments.

[0051] Unless otherwise specified, the information processing device performs any step in the embodiments of the present application, and the processor of the information processing device may perform the step. It is also worth noting that the embodiments of the present application do not limit the order in which the information processing device performs the following steps. In addition, the methods used to process data in different embodiments may be the same method or different methods. It should also be noted that any step in the embodiments of the present application can be independently executed by the information processing device, that is, when the information processing device executes any step in the following embodiments, it may not rely on the execution of other steps.

[0052] The present application embodiment provides an information processing method, which is applied to an information processing device, such as Figure 1 As shown, the method comprises the following steps:

[0053] Step 101: Obtain sample images, and train the sample images to generate a text detection model.

[0054] Among them, the sample picture is a picture with text information.

[0055] In the embodiment of the present application, the information processing device is any device with data processing capability.

[0056] It should be noted that the information processing device can use a machine learning algorithm to train sample images to generate a text detection model. The sample image is a picture containing but not limited to text information; at the same time, the font, color, font size and position of the text information in the sample image are not limited in any way in the embodiments of the present application; the format of the sample image can be a bitmap (Bitmap, BMP) format, a Joint Photographic Experts Group (Joint Photographic Experts Group, JPEG) format, a Portable Network Graphics (Portable Network Graphics, PNG) format, a Joint Photographic Experts Group (Joint Photographic Expert Group, JPG) format, etc., and the present application does not make any restrictions on this.

[0057] In the embodiment of the present application, the text information may be formed by any combination of words, letters, characters or numbers.

[0058] It should be noted that the text detection model is used to detect text information in sample images. At the same time, the text detection model can automatically binarize the image, that is, set the grayscale value of the pixel on the image to 0 or 255, that is, the image presents a visual effect of only black and white after binarization.

[0059] Step 102: Obtain sample information, and train the sample information to generate an information recognition model.

[0060] In the embodiment of the present application, the sample information may be formed by any combination of words, characters, numbers or letters; the information processing device generates an information recognition model by training the sample information using a machine learning algorithm.

[0061] Step 103: Obtain the image to be identified that is associated with the identification information.

[0062] In an embodiment of the present application, the identification information may be information that is input into the information processing device by a user in any manner and is recognized and acquired by the information processing device; wherein, the identification information may be any text information, and the present application does not impose any limitation on this.

[0063] In a feasible implementation, after the information processing device obtains the identification information "Happy New Year", it performs a web-wide crawler search for pictures to be identified that are associated with the identification information, ie, "Happy New Year".

[0064] In another feasible implementation, after the information processing device obtains the identification information of "final exam", it performs a network-wide crawler search for images to be identified that are associated with the identification information, ie, "final exam".

[0065] Among them, the image to be identified has text information that is somewhat related to the identification information; at the same time, the degree of association between the text information in the image to be identified and the identification information is not limited in this application.

[0066] Step 104: Identify the image to be identified based on the text detection model to obtain a text information set.

[0067] In an embodiment of the present application, the information processing device identifies and acquires the text information in the image to be identified through the text detection model generated by training, and obtains a text information set; wherein the text information set can be displayed in any form such as a table, a matrix, a graph, etc., and the present application does not impose any limitation on this.

[0068] It should be noted that the text information in the text information set may be sorted according to the number of characters in the text information, or according to the font size corresponding to the text information, or according to the attribute information of the to-be-identified image corresponding to the text information. This application does not impose any limitation on this.

[0069] Step 105: Process the text information set to determine a target vector.

[0070] In an embodiment of the present application, the text information in the text information set is processed according to preset rules to generate a target vector value corresponding to the text information set.

[0071] Step 106: Based on the information recognition model and the target vector, obtain the target text information associated with the identification information.

[0072] In an embodiment of the present application, the information processing device inputs the obtained target vector into a trained information recognition model, and processes it using a forward algorithm to obtain the target text associated with the identification information.

[0073] The information processing method provided by the embodiment of the present application obtains a sample image, and trains the sample image to generate a text detection model; wherein the sample image is a picture with text information; obtains sample information, and trains the sample information to generate an information recognition model; obtains a picture to be recognized associated with the identification information; recognizes the picture to be recognized based on the text detection model to obtain a text information set; processes the text information set to determine a target vector; obtains target text information associated with the identification information based on the information recognition model and the target vector; in this way, the picture to be recognized associated with the identification information is processed by the trained and generated text detection model to obtain a text information set, and the text information set is converted into a target vector, and the target vector is processed by the trained and generated information recognition model to obtain the target text information associated with the identification information, instead of using image recognition technology with cumbersome processing process and poor real-time performance to process network public opinion information with pictures as carriers as in the related art, thereby realizing the rapid extraction of text public opinion information in picture public opinion information.

[0074] Based on the above embodiments, the present application provides an information processing method, such as Figure 2 As shown, the method comprises the following steps:

[0075] Step 201: The information processing device obtains sample images and trains the sample images to generate a text detection model.

[0076] Among them, the sample picture is a picture with text information.

[0077] Step 202: The information processing device obtains sample information and trains the sample information to generate an information recognition model.

[0078] Step 203: The information processing device obtains a pre-processed image associated with the identification information.

[0079] In a feasible implementation, the information processing device may also implement step 203 by following steps 203a to 203b:

[0080] Step 203a: The information processing device obtains the image to be identified that is associated with the identification information.

[0081] The image to be identified is an image with multiple frames of static pictures.

[0082] In an embodiment of the present application, the image to be identified associated with the identification information is an image having multiple frames of static pictures; at the same time, the frames in the multiple frames of static pictures can be continuous or discontinuous; wherein, the image having multiple frames of continuous static pictures can be video information or an image in Graphics Interchange Format (GIF); a discontinuous multiple frames of static pictures can also be a GIF image.

[0083] Among the images of the multiple static picture frames, some of the static picture frames may have text information associated with the identification information, or each static picture frame may have text information associated with the identification information.

[0084] Step 203b: The information processing device performs analysis processing on the image to be recognized to obtain a pre-processed image.

[0085] In the embodiment of the present application, the information processing device processes the image to be identified through the data decomposition module and the similarity detection module in sequence to obtain a pre-processed image.

[0086] Step 204: The information processing device performs image enhancement on the pre-processed image to obtain the image to be identified.

[0087] In the embodiment of the present application, the information processing device may perform image enhancement on the preprocessed image in whole or in part to improve the visual effect of the preprocessed image.

[0088] It should be noted that image enhancement includes but is not limited to data enhancement of the pre-processed image and adjustment of the direction of the pre-processed image.

[0089] In a feasible implementation, the information processing device may also implement step 204 by executing the following steps 204a to 204c:

[0090] Step 204a: The information processing device obtains a sample image and trains the sample image to generate a direction detection model.

[0091] In an embodiment of the present application, the information processing device uses a machine learning algorithm to train sample images to generate a direction detection model; the information processing device can also generate a direction detection model based on a neural network pruning algorithm.

[0092] Among them, the direction detection model can divide the image information provided by the information processing device into four types of 0, 90, 180 and 270 according to preset rules, and then perform rotation operations based on the corresponding types to achieve rapid correction of the data direction of the sample image.

[0093] Step 204b: The information processing device uses a grayscale transformation algorithm to perform data enhancement on the preprocessed image to generate an enhanced image.

[0094] In an embodiment of the present application, the information processing device uses a grayscale transformation algorithm to perform data enhancement on the pre-processed image to improve the data quality of the pre-processed image.

[0095] The specific working principle of the grayscale transformation is as follows: for the grayscale range [a, b] of the preprocessed image f(x, y), it is hoped that the grayscale range of the enhanced image g(x, y) generated after the grayscale transformation is expanded to [c, d], which can be calculated by the following linear transformation formula (1).

[0096]

[0097] Step 204c: The information processing device determines the direction type of the enhanced image based on the direction detection model, and rotates the enhanced image based on the direction type and a preset rotation strategy to obtain a picture to be identified.

[0098] In an embodiment of the present application, the information processing device judges the direction of the enhanced image through the direction detection model generated by training, determines the four types corresponding to the enhanced image, namely the four types of 0, 90, 180 and 270 mentioned above, and then rotates the enhanced image according to the preset rotation strategy based on these four direction types to obtain a picture to be identified with a positive direction.

[0099] In a feasible implementation, the information processing device determines that the direction of the enhanced image is 90 degrees rotated clockwise, that is, type 90, based on the direction detection model; then the enhanced image is rotated 90 degrees counterclockwise according to the determined type 90 and the preset rotation strategy to obtain a picture to be identified with a positive direction.

[0100] Step 205: The information processing device recognizes the image to be recognized based on the text detection model to obtain a text information set.

[0101] Step 206: The information processing device processes the text information set to determine a target vector.

[0102] Step 207: The information processing device performs forward calculation on the target vector based on the information recognition model to obtain target text information.

[0103] In an embodiment of the present application, the information processing device inputs the target vector into a pre-trained information recognition model; wherein the information recognition model is generated by the information processing device through multiple rounds of training on sample information using a temporal classification loss function.

[0104] In an embodiment of the present application, the information processing device performs forward calculation on the target vector input to the information recognition model, that is, uses a forward algorithm to process it to obtain the target text information.

[0105] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.

[0106] The information processing method provided in the embodiment of the present application processes the image to be identified which is associated with the identification information and has multiple frames of static pictures through the trained text detection model to obtain a text information set, and converts the text information set into a target vector, and processes the target vector through the trained information recognition model to obtain the target text information associated with the identification information; thus, a more comprehensive acquisition of the image public opinion information and a rapid extraction of the text public opinion information in the image public opinion information are achieved.

[0107] Based on the above embodiments, the embodiments of the present application also provide an information processing method, such as Figure 3 As shown, the method comprises the following steps:

[0108] Step 301: The information processing device obtains sample images and trains the sample images to generate a text detection model.

[0109] Among them, the sample picture is a picture with text information.

[0110] Step 302: The information processing device obtains sample information and trains the sample information to generate an information recognition model.

[0111] Step 303, the information processing device obtains the image to be identified associated with the identification information.

[0112] The image to be identified is an image with multiple frames of static pictures.

[0113] Step 304: The information processing device analyzes the image to be recognized to obtain a pre-processed image.

[0114] In a feasible implementation, the information processing device may also implement step 304 by following steps 304a to 304b:

[0115] Step 304a: The information processing device decomposes the image to be recognized to obtain an intermediate image set.

[0116] The intermediate picture set includes at least two intermediate pictures.

[0117] In an embodiment of the present application, the information processing device can perform frame decomposition processing on the image to be identified, such as video information and GIF information, through an internal data decomposition module to obtain multiple frames of static images, i.e., an intermediate image set; wherein the image to be identified includes at least two frames of static images.

[0118] The number of static images obtained by decomposition is determined by the number of original frames of the image to be identified, namely, the video information and the GIF information.

[0119] It should be noted that the storage format of at least two intermediate pictures in the intermediate picture set is not limited in this application.

[0120] Step 304b: The information processing device performs similarity calculation on at least two intermediate images, and selects intermediate images from the intermediate image set to obtain pre-processed images.

[0121] In an embodiment of the present application, the information processing device converts each image vi to be identified into corresponding vector parameters according to their respective frame numbers, such as: v1: [a1, a2, a3, …, an], v2: [b1, b2, b3, …, bn], …, vi: [i1, i2, i3, …, in]; since there is a high degree of similarity between the static frame images in each image to be identified, direct identification processing by the information processing device will result in a waste of computing resources, thereby reducing the real-time performance of the processing process of the information processing device; and further, in the present application, the information processing device can perform similarity calculations on at least two intermediate images through an internal similarity detection module.

[0122] Among them, the information processing device performs similarity calculation mainly by calculating the multi-dimensional cosine distance between each vector parameter, and then based on a preset threshold, screens the intermediate images in the intermediate image set to obtain a pre-processed image.

[0123] In a feasible implementation, when the image to be identified having multiple frames of static images is video information, the video information is decomposed to obtain an intermediate image set including at least two intermediate images, and similarity calculation is performed on at least two intermediate images in the intermediate image set, and the intermediate images are screened from the intermediate image set to obtain a schematic diagram of four different pre-processed images.

[0124] Step 305: The information processing device performs image enhancement on the pre-processed image to obtain the image to be identified.

[0125] Step 306: The information processing device detects the image to be identified based on the text detection model to obtain text line information in the image to be identified.

[0126] In an embodiment of the present application, an information processing device performs text line detection on a picture to be identified by training a generated text detection model to obtain corresponding text line information; wherein the text detection model can be based on the following parameters: a feature pyramid network (FPN) structure and a Sigmoid function with coefficients to process the picture to be identified containing text information.

[0127] like Figure 4 As shown, it is a schematic diagram of a picture to be identified provided in an embodiment of the present application containing text line information, wherein the text line information included in the bold black line is such as: "third filter element", "high-quality coconut shell granular activated carbon", "remove odor and pigments in water", "pesticides and toxic substances", "adjust water quality to weak alkalinity" and "sweet taste suitable for human needs!", etc.

[0128] Step 307: The information processing device classifies the text line information to obtain a text information set.

[0129] In an embodiment of the present application, an information processing device classifies and processes text line information according to preset rules to obtain a text information set; wherein, the information processing device can classify the text line information according to attribute information such as the length or height of the text line information, and the present application does not impose any limitation on this.

[0130] Step 308: The information processing device obtains the length value and height value corresponding to each text line information in the text information set.

[0131] In the embodiment of the present application, the information processing device performs information recognition on each text line information in the text information set by using a text stacking algorithm; wherein the text line information in the text information set can be obtained by a text line set, such as: T = {a1, a2, a3, ..., a n} is stored and displayed, where n is the number of text lines in the text information set; at the same time, the information processing device obtains each text line information a in the text line set i The corresponding height value h i and length l i .

[0132] Step 309: The information processing device obtains a ratio value corresponding to each text line information based on the length value and height value corresponding to each text line information.

[0133] In the embodiment of the present application, the information processing device converts each text line information a i The corresponding height value h i and length l i By calculating through formula (2), we can get the information a of each text line: i The corresponding ratio value wi , where formula (2) is as follows:

[0134] w i =32 / (h i *l i ) (2)

[0135] Step 310: The information processing device obtains a target ratio value based on the ratio value corresponding to each text line information.

[0136] In the embodiment of the present application, the information processing device processes the acquired w i Sort to get the maximum value max_w i That is, the target ratio value; wherein the information processing device obtains w i Sorting can be done from largest to smallest or from smallest to largest.

[0137] Step 311: The information processing device constructs a vector based on the target ratio value to obtain a target vector.

[0138] In the embodiment of the present application, the information processing device is based on the maximum value max_w i To construct a four-dimensional vector value, that is, to generate a four-dimensional vector V T =(N, 1, 32, max_w i ); wherein the information processing device uses a slicing algorithm to sequentially i In descending order and through V T The corresponding vector is constructed in the form of , where the insufficient parts are filled in turn by adopting the zero assignment strategy to obtain the final target vector.

[0139] Step 312: The information processing device performs forward calculation on the target vector based on the information recognition model to obtain the target text information.

[0140] It should be noted that, for the description of the same steps and the same contents in this embodiment as those in other embodiments, reference can be made to the description in other embodiments and will not be repeated here.

[0141] The information processing method provided by the embodiment of the present application processes the image to be identified which is associated with the identification information and has multiple frames of static pictures through the text detection model generated by training to obtain a text information set, and converts the text information set into a target vector, and processes the target vector through the information recognition model generated by training to obtain the target text information associated with the identification information; thus, a more comprehensive acquisition of the image public opinion information and a rapid extraction of the text public opinion information in the image public opinion information are achieved.

[0142] Based on the foregoing embodiment, in the embodiment of the present application, when the information processing device obtains the target text information associated with the identification information based on the information recognition model and the target vector, that is, after the information processing device executes step 106, step 207 and step 312, it can also be implemented by the following steps A1 to A2:

[0143] Step A1: The information processing device constructs a sensitive corpus.

[0144] In an embodiment of the present application, an information processing device summarizes unhealthy or sensitive words on the Internet and in daily life to construct a sensitive corpus.

[0145] Step A2: The information processing device filters the target text information based on the sensitive corpus to obtain target security information.

[0146] In an embodiment of the present application, the information processing device can match the target text information with the content in the sensitive prediction library, that is, filter the target text information for sensitive words to remove bad words and sentences contained in the target text information to obtain the final target security information required.

[0147] like Figure 5As shown, it is a structural schematic diagram of a full-scale public opinion technology system processing provided in an embodiment of the present application; wherein, the acquired full-scale public opinion information including pictures, videos and dynamic pictures as carriers, that is, the image to be identified associated with the identification information in the embodiment of the present application, is processed by the following three modules: a preprocessing module, an information acquisition module and an information processing module, and the final public opinion information, that is, the target text information, is output; specifically, the preprocessing module includes a data decomposition module using data decomposition technology, a similarity detection module using a similarity detection algorithm to obtain frame pictures that meet the requirements in the input data, and a data enhancement module to improve the data quality. In the embodiment of the present application, the information processing device decomposes the image to be identified by the data decomposition module to obtain an intermediate picture set, and then calculates the similarity of at least two intermediate pictures in the intermediate picture set by the similarity detection module, thereby obtaining the preprocessed picture, and the similarity detection module. The preprocessed image is enhanced by a data enhancement module; wherein the data enhancement module adopts a grayscale transformation algorithm to achieve data enhancement of the preprocessed image; the information acquisition module includes a direction correction module, an information detection module and an information identification module that uses a deep learning algorithm to identify and obtain relevant information; the information processing module includes a sensitive information avoidance module, wherein the information processing device detects the direction type of the enhanced image mentioned in the embodiment of the present application through the direction correction module and rotates it according to a preset rotation strategy to obtain the image to be identified, and the direction correction module includes the direction detection model in the embodiment of the present application; the information detection module and the information identification module respectively include the text detection model and the information identification model in the embodiment of the present application; finally, the sensitive information avoidance module in the information processing module can filter the information to be filtered through the sensitive prediction library mentioned in the embodiment of the present application to achieve the purpose of sensitive information filtering.

[0148] Based on the above embodiments, the embodiments of the present application provide an information processing device 8, which can be applied to Figure 1-3 In an information processing method provided by a corresponding embodiment, referring to Figure 6 As shown, the information processing device 8 may include: a processor 81, a memory 82 and a communication bus 83, wherein:

[0149] The communication bus 83 is used to implement the communication connection between the processor 81 and the memory 82 .

[0150] The processor 81 is used to execute the program of the information processing method stored in the memory 82 to implement the following steps:

[0151] Obtain sample images, and train the sample images to generate a text detection model; wherein the sample images are images with text information;

[0152] Obtain sample information, and train the sample information to generate an information recognition model;

[0153] Obtain the image to be identified that is associated with the identification information;

[0154] Based on the text detection model, the image to be identified is identified to obtain a text information set;

[0155] Processing the text information set to determine the target vector;

[0156] Based on the information recognition model and the target vector, the target text information associated with the identification information is obtained.

[0157] In other embodiments of the present application, the processor 81 is used to execute the steps of acquiring the image to be identified associated with the identification information stored in the memory 82 to implement the following steps:

[0158] Obtaining a preprocessed image associated with identification information;

[0159] Perform image enhancement on the preprocessed image to obtain the image to be recognized.

[0160] In other embodiments of the present application, the processor 81 is used to execute the acquisition of the pre-processed image associated with the identification information stored in the memory 82 to implement the following steps:

[0161] Acquire an image to be identified that is associated with the identification information; wherein the image to be identified is an image having multiple frames of static pictures;

[0162] The image to be recognized is analyzed and processed to obtain a preprocessed image.

[0163] In other embodiments of the present application, the processor 81 is used to execute the analysis processing of the image to be identified stored in the memory 82 to obtain a pre-processed image, so as to implement the following steps:

[0164] Decomposing the image to be recognized to obtain an intermediate picture set; wherein the intermediate picture set includes at least two intermediate pictures;

[0165] The similarity of at least two intermediate images is calculated, and the intermediate images are selected from the intermediate image set to obtain preprocessed images.

[0166] In other embodiments of the present application, the processor 81 is used to execute the image enhancement of the pre-processed image stored in the memory 82 to obtain the image to be identified, so as to implement the following steps:

[0167] Obtain sample images, and train the sample images to generate a direction detection model;

[0168] The grayscale transformation algorithm is used to perform data enhancement on the preprocessed image to generate an enhanced image;

[0169] The direction type of the enhanced image is determined based on the direction detection model, and the enhanced image is rotated based on the direction type and a preset rotation strategy to obtain a picture to be identified.

[0170] In other embodiments of the present application, the processor 81 is used to execute the text detection model stored in the memory 82 to identify the image to be identified, and obtain a text information set to implement the following steps:

[0171] Detect the image to be identified based on the text detection model to obtain the text line information in the image to be identified;

[0172] The text line information is classified to obtain a text information set.

[0173] In other embodiments of the present application, the processor 81 is used to execute the steps stored in the memory 82 to process the text information set and determine the target vector, so as to implement the following steps:

[0174] Get the length and height values ​​corresponding to each text line in the text information set;

[0175] Based on the length value and height value corresponding to each text line information, obtain the scale value corresponding to each text line information;

[0176] Based on the ratio value corresponding to each text line information, obtain the target ratio value;

[0177] A vector is constructed based on the target ratio value to obtain a target vector.

[0178] In other embodiments of the present application, the processor 81 is used to execute the information recognition model and the target vector stored in the memory 82 to obtain the target text information associated with the identification information, so as to implement the following steps:

[0179] The target vector is forward calculated based on the information recognition model to obtain the target text information.

[0180] In other embodiments of the present application, the processor 81 is used to execute the information recognition model and the target vector stored in the memory 82, and after obtaining the target text information associated with the identification information, the following steps may be implemented:

[0181] Build a sensitive corpus;

[0182] The target text information is filtered based on the sensitive corpus to obtain the target security information.

[0183] It should be noted that the specific implementation process of the steps executed by the processor in this embodiment can refer to Figure 1-3 The implementation process of the information processing method provided in the corresponding embodiment will not be repeated here.

[0184] The information processing device provided by the embodiment of the present application processes the image to be identified that is associated with the identification information and has multiple frames of static pictures through the text detection model generated by training to obtain a text information set, and converts the text information set into a target vector, and processes the target vector through the information recognition model generated by training to obtain the target text information that is associated with the identification information and removes sensitive words; thereby achieving more comprehensive acquisition of picture public opinion information and rapid extraction of text public opinion information that removes sensitive words from the picture public opinion information.

[0185] Based on the foregoing embodiments, the embodiments of the present application provide a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the following. Figure 1-3 The corresponding embodiments provide steps of the information processing method.

[0186] It should be noted that the processor or CPU may be at least one of an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a Central Processing Unit (CPU), a controller, a microcontroller, and a microprocessor. It is understandable that the electronic device that implements the above-mentioned processor function may also be other, and the embodiments of the present application are not specifically limited.

[0187] It should be noted that the above-mentioned computer storage medium / memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM) and other memories; it can also be various terminals including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.

[0188] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0189] In addition, all functional units in the embodiments of the present application can be integrated into one processing module, or each unit can be a separate unit, or two or more units can be integrated into one unit; the above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units. A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium, which, when executed, executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memory (ROM), random access memory (RAM), disks or optical disks, and other media that can store program codes.

[0190] The methods disclosed in several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0191] The features disclosed in several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0192] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An information processing method, characterized in that: The method comprises: Obtaining a sample image, and training the sample image to generate a text detection model; wherein the sample image is an image with text information; Acquire sample information, and train the sample information to generate an information recognition model; Obtain the image to be identified that is associated with the identification information; Identify the image to be identified based on the text detection model to obtain a text information set; Obtaining a length value and a height value corresponding to each text line information in the text information set; obtaining a ratio value corresponding to each text line information based on the length value and the height value corresponding to each text line information; obtaining a target ratio value based on the ratio value corresponding to each text line information; constructing a vector based on the target ratio value to obtain a target vector; Based on the information recognition model and the target vector, target text information associated with the identification information is obtained.

2. The method according to claim 1, characterized in that The step of obtaining the image to be identified that is associated with the identification information includes: Obtaining a preprocessed image associated with the identification information; Perform image enhancement on the preprocessed image to obtain the image to be identified.

3. The method according to claim 2, characterized in that The step of obtaining a pre-processed image associated with the identification information includes: Acquire an image to be identified that is associated with the identification information; wherein the image to be identified is an image having multiple frames of static pictures; The image to be identified is analyzed to obtain the pre-processed image.

4. The method according to claim 3, characterized in that The step of parsing the image to be identified to obtain the pre-processed image includes: Decomposing the image to be recognized to obtain an intermediate picture set; wherein the intermediate picture set includes at least two intermediate pictures; Similarity calculation is performed on the at least two intermediate pictures, and intermediate pictures are selected from the intermediate picture set to obtain the pre-processed picture.

5. The method according to claim 2, characterized in that: The performing image enhancement on the pre-processed image to obtain the image to be identified includes: Acquire a sample image, and train the sample image to generate a direction detection model; Performing data enhancement on the preprocessed image using a grayscale transformation algorithm to generate an enhanced image; The direction type of the enhanced image is determined based on the direction detection model, and the enhanced image is rotated based on the direction type and a preset rotation strategy to obtain the image to be identified.

6. The method according to claim 1, characterized in that The step of identifying the image to be identified based on the text detection model to obtain a text information set includes: Detect the image to be identified based on the text detection model to obtain text line information in the image to be identified; The text line information is classified to obtain the text information set.

7. The method according to claim 1, characterized in that The obtaining, based on the information recognition model and the target vector, target text information associated with the identification information includes: The target vector is forward calculated based on the information recognition model to obtain the target text information.

8. The method according to claim 1 or 7, characterized in that: After obtaining the target text information associated with the identification information based on the information recognition model and the target vector, the method further includes: Build a sensitive corpus; The target text information is filtered based on the sensitive corpus to obtain target security information.

9. An information processing device, characterized in that: The information processing device comprises: a processor, a memory and a communication bus; The communication bus is used to realize the communication connection between the processor and the memory; The processor is used to execute the program in the memory to implement the information processing method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of the information processing method according to any one of claims 1 to 8.

Images