49 results about "Base64" patented technology

The invention discloses a Web interface element recognition method, system and device and a medium, belongs to the technical field of artificial intelligence and Web interface development, and aims to solve the technical problem of how to improve the recognition rate of dynamic elements and CSS hidden elements so as to enable Web interface element positioning to be more accurate. Performing screenshot by using a playwright screenshot function, converting the screenshot into a Base64 coded character string by using a function after the screenshot is completed, inputting the character string and a url into an intelligent agent, calling a DeepSeek-VL2 visual model by the intelligent agent, analyzing the screenshot and outputting identified structured element data; dOM structure analysis: inputting a page url address, and generating DOM structure data by using a DOM analysis function; dOM tree analysis and visual element matching: calling an element matching function, performing visual element and DOM structure matching, and obtaining all successfully matched element data; interface element extraction and positioning generation; and element application.

The invention discloses a KNN retrieval method of a remote sensing image in a visual language model, which comprises the following five steps: S1, data preprocessing: performing base64 processing on a remote sensing image data set in a visual language test data set, performing jsonl processing on a text data set, and uploading the processed remote sensing image data set to a database; s2, serializing the image text: serializing the preprocessed data set to facilitate random reading during training; s3, CLIP model training: sending the serialized data into a visual language model for training to obtain a training model; s4, image-text feature extraction: performing feature extraction on the test data through the training model; and S5, KNN retrieval evaluation: carrying out KNN retrieval evaluation on the file subjected to feature extraction. According to the method, the KNN retrieval can be trained through the visual language test data set, and retrieval evaluation is carried out, so that relevant or similar contents can be quickly and accurately retrieved from mass data.

The invention discloses an RAG document splitting optimization method and system in the operator field, and belongs to the technical field of large model optimizing.The method comprises the steps that a document is uploaded, and source files and file information are stored through minio and paradeDB respectively; constructing a document loader, automatically selecting a corresponding loader according to a file type to carry out file processing, converting the file into a uniform Markdown text, analyzing a picture in the document, and converting the picture into a base64 format; constructing a picture processor, extracting a picture base64 character string in the text, processing a bitmap and a vector diagram, and converting the bitmap and the vector diagram into a Markdown picture reference format; constructing a document divider; vector conversion; and recalling the text. According to the method, the problem that picture extraction, conversion and processing are complex in the current RAG is solved, vector retrieval, full-text retrieval and mixed retrieval functions can be achieved through one database, and the operation and maintenance cost is reduced.

The invention provides an e-commerce commodity graph differentiation generation method and device, equipment and a medium, and the method comprises the steps: obtaining an original to-be-shelved commodity graph of e-commerce, and finally obtaining the disassembly information; the method comprises the following steps: adjusting an original commodity image into 980 * 980 pixels, performing Gaussian blur processing at the same time, and then converting image pixel data into a Base64 coded character string; constructing a composite cue word according to the image data, the disassembly information and a format required by generation; inputting the composite cue word into a model to generate a new scene description; inputting the new scene description, the original commodity graph and the synthesis requirement into a model to generate a synthesis cue word, and then inputting the original commodity and the synthesis cue word into an image-text model to generate a differentiated commodity graph; setting the output resolution, the number of iterations and the learning rate of the SuperIR model, and inputting the differentiated commodity graph into the SuperIR model to generate a differentiated commodity optimization graph; the defects of low efficiency, poor scene adaptation, main body distortion and difficulty in repeated avoidance of a traditional method are overcome.

The application discloses a QR code encryption and decryption method and system based on chaos theory and an AES algorithm, relates to the encryption and decryption technical field of QR codes, and comprises an encryption process and a decryption process; the encryption process comprises the following steps: firstly, chaotic sequences generated by improved Logistic mapping and Chebyshev mapping are subjected to XOR to obtain a composite chaotic sequence, the composite chaotic sequence is used as an initial key of AES grouping, and then ciphertext is obtained through AES encryption. Grouping initial keys are dynamically generated, a newly generated key is used for encrypting one group of plaintext data, the keys are completely independent, and therefore, one-time password is realized. Then, the AES algorithm is improved, plaintext information is first encrypted, then secondary encryption is performed on the plaintext information in a Base64 mode, final ciphertext information is subjected to QR code standard coding, and an encrypted QR code is generated. Through the mode of increasing encryption processing, the application effectively reduces information leakage or information tampering problems caused by QR codes.

The invention relates to the technical field of cross-system data processing, and particularly provides an intelligent clipboard content processing method and system, and the method comprises the following steps: S1, content type detection; s2, Base64 decoding is carried out; s3, performing path conversion; and S4, output control. Compared with the prior art, intelligent conversion and standardized output of clipboard content can be realized through dynamic content identification and multi-format output control.

The present invention relates to a GIS and BIM fusion modeling system for urban intelligent planning. The system comprises a data processing module, a model storage module, and a model processing module. The data processing module obtains GIS data and BIM data, fuses them, and generates a fusion model. The fusion model is stored in a database via the model storage module and is called by the model processing module. BIM data formats include wexbim and IFC. BIM data in wexbim format is converted to Base64 before storage. BIM data in IFC format is first converted to wexbim format and then to Base64 before storage. Compared with the prior art, the present invention has the advantages of increasing the flexibility of BIM data processing and storage, effectively alleviating the difficulties in processing and storing multivariate data in urban intelligent planning, and improving the modeling integrity of the entire area.

The invention belongs to the technical field of computers, and provides a pressure testing method, device and system and a computer readable storage medium, and the method comprises the steps: firstly, a tested end receives a current HTTP request sent by a testing front end, and obtains a coding request body of the current HTTP request; the encoding request body is encrypted base64 encoding data; and then the tested end calls a user-defined plug-in to sequentially carry out decoding, JSON parameterization processing and secondary coding processing on the coding request body to obtain a target request body subjected to secondary coding, and carries out pressure testing on the target request body, so that the performance testing capability can be improved, and the resource consumption of a pressing machine is reduced.

The invention discloses a method, a device, a medium and equipment for a browser client to play RTSP (Real Time Streaming Protocol) video stream data. The invention belongs to the technical field of computers. The method comprises the following steps: acquiring configuration information of a front-end configuration interface on a camera through a preset interface; pulling an RTSP video stream from the camera according to the configuration information; capturing the RTSP video stream frame by frame, and obtaining compressed data based on an inter-frame compression technology; performing Base64 coding conversion on the compressed data to obtain coded data in a picture format; and pushing the coded data to a video playing device so as to dynamically display the coded data through a browser client of the video playing device. By adopting the technical scheme, the 1t can be used in the browser; imgt, imgt; according to the method for playing the RTSP video through the browser and the tag, the direct support limitation of the browser on the RTSP video stream is ingeniously bypassed, additional plug-ins are not needed, and a high-efficiency, low-delay and low-cost real-time RTSP video stream data playing scheme is achieved.

The invention discloses a high-performance asynchronous segmentation Base64 coding method and device based on a VTE system. The method comprises the following steps: dynamically adjusting the segmentation size according to the data type; the method comprises the following steps: uniformly converting an Array Buffer and a character string into a character string coded by UTF-8 (Untranslated Transform Factor-8) by adopting a preprocessor Data method, and processing illegal characters according to a fault-tolerant strategy at the same time; the method comprises the following steps of: completing single-segment coding by adopting an encoder Segment, and putting coding logic into a micro-task queue through await; calculating and displaying the encoding progress; a yieldMain Thread method is adopted, and a main thread is preferentially made to process key operations or key tasks; therefore, large file coding is prevented from blocking a main thread, and real-time response of key operations such as image preview and medical record editing of the VTE system is guaranteed; unicode medical characters are adapted, illegal character fault tolerance of medical data is supported, it is ensured that the coding process is not interrupted, and complete transmission of the data is guaranteed; medical scene progress feedback is provided, the operation experience of medical staff is improved, and repeated operation is avoided; the method is compatible with a native btoa API, and the transformation and migration cost of a VTE system is reduced.

The invention discloses an autonomous learning method in visual saliency target detection, which comprises the following steps of: (1) receiving electroencephalogram data in real time through a data receiving thread, processing and analyzing information such as a task ID (Identity), a patient ID and electroencephalogram data in the electroencephalogram data, and storing the information in a dynamically named data dictionary to wait for being called by a drawing thread; (2) after a drawing thread detects that a new dynamic dictionary is generated in real time, the data are immediately sent into an electroencephalogram data analysis module for operations such as wave trapping and filtering, and the average power spectral density is calculated after short-time Fourier transform is carried out; and (3) after the average power spectral density is calculated in the step (2), the rendering of each point in the high-density two-dimensional network is calculated and created by adopting a spherical interpolation method according to created channel information and the like, and finally, a base64 character string is generated by each picture. The drawing thread returns to the front end for display in real time, and the display effect of the real-time brain topographic map is formed.

The invention discloses a method and a system for stably displaying a Base64 picture at a Web end. The server performs Base64 coding on pictures not exceeding 1MB, generates content fingerprints, associates versions with metadata, and distributes description data through homologous signatures; and the front end asynchronously loads and displays after coding and optional zooming on the canvas side, and when the loading fails, hierarchical backspacing taking the placeholder as the first is executed. The system can selectively start a fragmented list and memory water level control, combines local cache and flicker-free replacement with fingerprints and versions as keys, and supports operation rapid picture changing and history tracing. According to the scheme, the problems of cross-domain pollution, first screen blank and memory peak are solved, and the loading stability and the operation and maintenance efficiency are improved.

The invention discloses an unstructured data processing method and device. The method comprises the following steps: establishing a secure data channel between a bank external system and a bank internal unstructured data storage system; based on the security data channel, receiving an access request from a bank external system; processing unstructured data according to the type of the access request; when the access request is a query request, generating an unstructured data identifier based on a request parameter; calling a software toolkit of the unstructured data storage system to pull a corresponding file of the query request to local storage, and forming an unstructured data file and a corresponding return virtual path; receiving an access instruction for the unstructured data file; and when the access request is an uploading request, decoding the Base64 coding file, calling the software toolkit, and storing the software toolkit in an unstructured data storage system. The method and the device are used for improving the efficiency and the accuracy of querying, uploading and maintaining the non-structural data.

This invention discloses an asynchronous segmented Base64 decoding method and system for a hospital-wide integrated clinical rehabilitation system. The method includes: S1: Dynamic segmentation; parsing the Base64 encoded string to identify the data type, and dividing the segment length according to the data type and the total data length; S2: Asynchronous task scheduling; dividing the decoding task into multiple levels according to the priority of medical business, and allocating the decoding task to the core thread pool and elastic thread pool according to the level; S3: Parallel decoding; precompiling the mapping relationship between Base64 encoded characters and their corresponding binary values ​​into a hash table, and using SIMD instructions to perform parallel operations on the segmented encoded data; S4: Fault tolerance verification; verifying integrity; S5: Result merging and callback; thereby improving decoding efficiency and ensuring integrity.

The invention discloses a Base64 invoice automatic read-write optimization method based on artificial intelligence. The method comprises the following steps: S1, executing integrity verification on an electronic invoice file uploaded by a user and generating Base64 coded invoice data; s2, transmitting the Base64 coded invoice data to a cloud service platform through an HTTPS protocol, and executing field structured extraction processing; s3, calling an invoice data interface of the tax system, and performing validity and consistency check on the structured invoice field set; s4, inputting the structured invoice field set into the evidence deep learning model to generate an uncertainty analysis result; s5, respectively calculating a data uncertainty score and a model uncertainty score; s6, judging the relationship between the uncertainty decomposition result and a preset threshold value; and S7, if the uncertainty decomposition results do not exceed the corresponding preset threshold value, marking the structured invoice field set and the tax check result as a credible state. According to the invention, an evidence deep learning model is adopted, and automatic Base64 invoice reading and writing are realized.

A suspicious Uniform Resource Locator (URL) of a resource on the public Internet is detected by searching a part of the URL for a candidate substring, which is the longest possible substring that is in accordance with a Base64 encoding scheme. The candidate substring is converted to a candidate binary data in accordance with the Base64 encoding scheme. The candidate binary data is then converted to a candidate American Standard Code for Information Interchange (ASCII) string. The candidate ASCII string is evaluated to determine the information type of the candidate ASCII string. A determination as to whether the URL is suspicious is based at least on the information type of the candidate ASCII string.

The invention discloses a flattening HTTP (Hyper Text Transport Protocol) decoding method and system based on multi-classification learning and a medium. A multi-classification HTTP coding detection model is generated by constructing an HTTP coding data set, extracting a multi-classification feature matrix of the coding data set and training the coding data set through multi-classification learning; extracting an HTTP original load from the HTTP traffic, extracting a multi-classification feature vector of the HTTP original load, loading a multi-classification HTTP coding detection model, and carrying out multi-classification HTTP coding detection; judging an HTTP coding detection result; if the code of the original load in the HTTP flow is detected, flattening HTTP decoding is carried out; recording and outputting an HTTP decoding result; according to the method, the original load is extracted from the HTTP flow, multi-classification machine learning training detection model is carried out aiming at multiple types of codes such as URL, Base64, Unicode, Hex and HTML, flattening decoding operation is carried out according to a model detection result, and finally the clear HTTP load is cleaned out.

The invention discloses a multi-type HTTP (Hyper Text Transport Protocol) coding feature extraction method and system and a medium. The method comprises the following steps: constructing a multi-type HTTP coding data set; sample data of URL codes, Base64 codes, Unicode codes, Hex codes, HTML codes and combined codes and nested codes of the URL codes, the Base64 codes, the Unicode codes, the Hex codes and the HTML codes are collected respectively, and corresponding type labels are marked; the method comprises the following steps: reading a multi-type HTTP coded data set, sequentially reading samples, and carrying out feature extraction; constructing a multi-type coding feature matrix; constructing a coding sample feature vector, constructing and standardizing a feature matrix of the multi-type HTTP coding data set, constructing a coding type vector matrix according to type values in the multi-type HTTP coding data set, and performing transverse splicing operation on the coding type vector matrix and the standardized feature matrix to construct a multi-type coding feature matrix; according to the method, a multi-type coding data set which can be used for coding learning training, verification and testing is constructed, feature extraction is performed through unique features of multi-type coding, and the accuracy and efficiency of feature extraction are improved.

The invention discloses a dynamic identity authentication method and system based on an HMAC-SHA256 algorithm and a medium, and mainly relates to the technical field of identity authentication. The method and the device are used for solving the problems that in an existing dynamic identity authentication scheme, a static key is prone to brute force cracking or key leakage, a fixedly arranged message body structure is cracked, and a receiving end lacks an effective time window verification mechanism. Comprising the following steps: encrypting a message body by using a data key through an HMAC-SHA256 algorithm, then carrying out Base64 transcoding to obtain a final message body, and sending the final message body to a receiver; when a receiver receives a final message body, decoding the final message body by using Base64, and obtaining the message body by using a data key; disassembling the message body by using a preset permutation and combination mode and a preset separator; and when the server timestamp in the message body is within a preset synchronization interval, identity authentication code verification is carried out to receive a message plaintext after verification succeeds.

The invention discloses a cloud desktop real-time auditing method based on an RDP protocol, and aims to solve the problems that an existing offline auditing method cannot perceive and intervene violation operation of a user in real time and excessively depends on human resources. The method comprises the following steps of: firstly, assigning connection parameters of a target host and connecting through an agent; secondly, acquiring a base64 code of a current operation fragment graph of a user from an RDP (Remote Decision Protocol) channel and carrying out user information modeling; then, decoding the base64 code into a picture, judging whether the resolution of the picture needs to be improved or not according to administrator configuration or machine load conditions, and identifying character information in the picture; and finally, matching the user behavior character information with a preset auditing rule, and alarming an administrator if the user behavior character information is hit. According to the method, the characteristic of transmitting the fragment graph through the RDP is utilized, capturing, processing and auditing of user operation behaviors are achieved, the timeliness and flexibility of auditing are improved, manpower dependence is reduced, and the defect of off-line auditing is effectively overcome.

The invention provides a BASE64-based message data encryption method, a BASE64-based message data decryption method and related products, and the encryption method comprises the steps: carrying out BASE64 coding on first message data to obtain a first coded character string; determining a first offset number according to the character string length of the first coded character string, and determining a first index value set corresponding to the first coded character string according to the first offset number; the first index value set comprises a plurality of first index values; querying a preset mapping character table, and determining a first encryption character corresponding to each first index value in the first index value set; and generating a first encrypted character string for the first message data according to the first encrypted character. According to the embodiment of the invention, the security of message data encryption can be improved. And additional codes and identifiers are not generated, and any additional secret key is not used. In addition, the coding mode is simple, the calculated amount is small, system resources can be saved, and therefore performance consumption is reduced.

The application discloses an intelligent canvas interaction system and method based on AIAgent. The system comprises a front-end canvas module, a ChatWebSocket channel, a BridgeWebSocket channel, an AIAgent decision module, a tool registration and routing module, and a front-end tool executor. The system assigns an execution position identifier to each tool in the tool set and defines a routing decision function. The canvas state is synchronized using an incremental superposition mechanism, and the layer set, element set, and configuration parameters are updated according to a formula after the first tool call. The system supports the canvas_state tool for AI to obtain the current canvas state in real time to assist in parameter decision-making, and the canvas_preview tool for performing a screenshot on the canvas and returning a base64 encoded image for AI visual verification. The method comprises the steps of instruction receiving, semantic analysis, tool routing, cross-end execution, result summarization, and response generation. The application realizes natural language driven multi-step canvas operation, effectively reducing the user threshold.

The invention discloses an implementation method of a text field component based on Base64 coding, and relates to the field of cloud computing and web application. According to the method, a Base64 coding-based text field component is created, the text field component comprises two modes, the text field component comprises a single-choice box for displaying Base64 coded content and a component for text field control extension in the text mode, a text is input in the text mode to generate text data in a Base64 coding format, and the text data is stored in the single-choice box. The text field component comprises an uploading button and a text field control extension component in the picture mode, and a Base64 format character string is generated from the selected picture in the picture mode.

The application provides a kind of compatible DBC's automobile communication method and system in the field of automobile communication, and the method comprises the following steps: step S10, based on DBC format definition DBCE format;Step S20, when automobile communication, the sending end is based on the DBCE format and the CAN message to be transmitted generation DBCE file, and the DBCE file is sent to receiving end after being encrypted using Base64 algorithm;Step S30, the receiving end is decrypted after receiving the DBCE file using Base64 algorithm, and based on the DBCE format, DBCE file is parsed in real time, and then automobile communication is completed.The application has the advantages that: the flexibility, expansibility, efficiency and safety of automobile communication are greatly improved.

The present invention provides a system (100) and method (200) for embedding custom metadata in Over-The-Top (OTT) streaming manifests to enhance the user experience. The system (100) includes an encoder and packager module (108) configured to generate Adaptive Bitrate (ABR) segments and corresponding media manifests. A manifest updater (112) embeds custom metadata within the manifests, including playback control information such as content advisories, scene-specific annotations, and interactive elements. The custom metadata is embedded as XML elements for DASH and UTF-8 tags for HLS, encoded in base64-encoded JSON format. The media manifests with embedded metadata are transmitted to a Content Delivery Network (CDN) (104), which delivers them to client devices (106). The client devices (106) may process the custom metadata or continue streaming without interruption if the metadata is not recognized, ensuring backward compatibility. The method (200) facilitates seamless integration of metadata, enhancing playback without affecting legacy systems.

The application discloses a picture transmission method and device, a picture display method and device, and relates to the technical field of computers.The picture transmission method comprises the following steps: obtaining picture data, a picture encryption key, a transmission encryption key and encrypted interference data; encrypting the picture data by using the picture encryption key to obtain a picture encryption result, inserting the picture encryption key into the picture encryption result to obtain an insertion result, and recording insertion position data; performing base64 encoding on the insertion position data and the encrypted interference data to obtain a processing result; splicing the insertion result and the processing result, encrypting the spliced result by using the transmission encryption key to obtain transmission data; performing base64 encoding on the encryption algorithm name information of the encryption key to obtain an encoding result; and sending the encoding result and the transmission data to a receiving end.In the application, the picture data does not directly participate in data transmission, and the security and privacy of the picture data in the picture transmission process are improved.

The invention provides a Java file hybrid encryption method based on dynamic blocking and compression, and relates to the technical field of computer security, and the method comprises the following steps: generating an RSA key pair, and encrypting and storing the key through Base64 coding; the method comprises the following steps: converting a Java file into binary data, dividing the binary data into a plurality of blocks according to RSA encryption length limitation, respectively carrying out Base64 coding and RSA encryption, generating encrypted blocks, integrating the encrypted blocks into an encrypted file, and generating a digital signature by using a private key to ensure the integrity. In the loading process, automatic decryption and signature verification are conducted through the user-defined class loader embedded with decryption logic, all encryption and decryption operations are achieved through a C + + dynamic link library, interaction with a Java program is achieved through a JNI interface, and safety and execution efficiency are improved.

The invention provides a fabric edge database augmentation labeling method for deep learning. The method comprises the following steps: acquiring a surface image of the composite material woven fabric through an industrial camera, labeling by using LABELME software, and generating a corresponding JSON labeling file; augmenting the annotated image, such as horizontal mirroring, vertical mirroring, rotation, graying, sharpening and the like, and correspondingly transforming the annotated information in the JSON annotated file to match the augmented image; and converting the augmented image into a Base64 code, and updating the image information in the JSON annotation file. The method has the characteristics of intelligence, high efficiency and accuracy, the accuracy and universality of deep learning in fabric edge automatic measurement can be remarkably improved, and the database construction cost is effectively reduced.

The invention discloses a Keyblock user name password encryption transmission method and device and a medium, and the method comprises the steps: injecting an RSA public key character string into a front-end login page, encrypting a plaintext password inputted by a user before a form is submitted, and generating a Base64 coded ciphertext; replacing the password field value in the original form with the encrypted ciphertext, and submitting the form carrying the password of the ciphertext to a Keyblock server; the method comprises the following steps: at a Keyblock server side, acquiring a ciphertext password through a self-defined Authority, and decrypting by using a pre-stored RSA private key to obtain a corresponding plaintext password; and re-injecting the decrypted plaintext password into the Keyblock standard authentication process to complete the verification of the user name password. The full-link encryption from the client to the server is ensured, and the risk of application layer plaintext leakage which cannot be protected by the HTTPS protocol is solved.

The invention provides an e-commerce multi-grid commodity graph differentiation generation method and device, equipment and a medium, and the method comprises the steps: obtaining an original e-commerce multi-grid commodity graph, extracting each sub-graph, and recognizing the commodity main body type of each sub-graph; finally, disassembling information is obtained; the size of each sub-image is changed into 980 * 980 pixels, noise is removed, and the sub-images are converted into Base64 codes; constructing a corresponding composite prompt word according to the image data, the disassembly information and a format required by generation; inputting each composite cue word into a model to generate a new scene description corresponding to each sub-graph; inputting the original multi-grid commodity graph, the new scene description corresponding to each sub-graph and the set constraint into a model to generate a replacement prompt word; inputting the replacement prompt word and the original multi-grid commodity graph into a model to generate a multi-grid commodity replacement graph; the defects of scene replacement generalization, complex positioning, style splitting and low efficiency of a traditional method are overcome.