84 results about "Glyph" patented technology

This invention relates to the field of text processing technology, and in particular to a method, device, and medium for processing navigational notice text based on semantic enhancement. The method includes: first, acquiring a content carrier to be processed; then, acquiring the semantic vector and glyph feature vector of the content carrier; concatenating the two types of vectors to form an enhanced text representation; extracting temporal features from the enhanced text representation to obtain temporal features containing forward and backward logical relationships within the text; acquiring the weights of words and sentences in the temporal features and performing weighting to obtain weighted word representations and weighted sentence representations; performing correction processing on the weighted representations to generate corrected text; and finally, validating the corrected text and outputting the target text. This invention can improve the accuracy and efficiency of content carrier processing.

The invention discloses a stylized visual text editing method, a stylized visual text editing system, stylized visual text editing equipment and a storage medium, which are corresponding schemes, and the related schemes aim to solve the problem of style consistency existing in image text editing of an existing diffusion model, and the stylized visual text editing efficiency is improved by combining visual features of a font image and an input text image. The method comprises the following steps of: extracting style embedded information from a text image, and inputting the style embedded information as an enhanced style condition into a diffusion model to realize fine control on a diffusion process, so that the diffusion model can generate a text image with high readability and style consistency, and can realize maintenance of an original text style or style migration based on a reference image.

The invention provides a teaching document anti-copy protection method and system based on OCR interference resistance, and belongs to the technical field of information security and digital copyright protection. According to the method, anti-OCR font processing is applied to document characters, invisible disturbance symbols are inserted between characters or paragraphs, watermark information is embedded in document typesetting or font strokes, and finally a standard PDF or DOCX file is compiled. Watermark information is embedded through stroke thickness change, paragraph row spacing indentation perturbation and other modes. The system comprises a font disturbance module, a character disturbance module, a watermark embedding module, a compiling output module and a verification module. According to the method, the human eyes can normally read the document, but messy codes or disorder can occur when the OCR tool extracts the document, and the document cannot be directly reused; meanwhile, watermark information cannot be stripped, so that leakage tracking is facilitated; and the output file keeps a standard format and is compatible with the existing office teaching process. The method is suitable for anti-copying protection of teaching documents, scientific research data, test papers, contract files and the like.

The invention relates to the technical field of content extraction, in particular to a multi-modal hybrid document OCR (Optical Character Recognition) and structured extraction method, which comprises the following steps of: acquiring an image text region bounding box and classifying a style, extracting a font or stroke sequence to generate a character positioning structure, dividing paragraph and sentence groups to classify semantic fields, and calculating a field matching relationship to generate structural mapping. According to the method, logic mapping is constructed through character two-dimensional coordinate sorting and paragraph contours, complete reconstruction of a page structure is enhanced, semantic field categories are extracted by using syntactic density of sentence paragraph division and inter-paragraph features, the accuracy of field classification is improved, and the method has the advantages of being simple in structure, convenient to operate and high in practicability. A field mapping relation is established through Jaccard similarity and part-of-speech consistency analysis between a head word and a standard field keyword, a field path index and a structure node link are clarified, field semantic affiliation and structure position output are unified, and the document structure reduction degree and field extraction accuracy are improved.

The invention discloses an interactive Chinese character traceability and writing correction system, and belongs to the technical field of Chinese character education. Aiming at the problems of traceability staticization, correction simplification and path solidification of an existing tool, the technical scheme comprises the following steps: constructing a database of seven types of font evolution from oracle to regular script, and dynamically displaying font evolution by combining 3D modeling and AR rendering; multi-mode writing data are collected through a pressure sensing pen, an infrared camera and the like, and the AI model is corrected in real time from pressure, posture and structure dimensions; and on the basis of the user error type, related Chinese character traceability content is pushed by reinforcement learning and stepped practice is carried out, so that personalized path optimization is realized. The system can improve Chinese character meaning understanding depth, writing normalization and learning efficiency, is adaptive to scenes of basic education, calligraphy training and the like, and promotes Chinese character learning to be converted from passive cognition to active construction.

The application relates to the technical field of natural language, and discloses a text error correction processing method, which comprises the following steps: obtaining a text to be corrected; performing feature extraction on the text to be corrected through a feature layer of a text error correction model to obtain position features, image features and font features; performing error labeling on text characters in the text to be corrected according to the position features, the image features and the font features through a labeling layer of the text error correction model to obtain error characters; performing stroke decomposition on each error character to obtain character strokes; performing character matching on the character strokes of each error character through a preset homophone dictionary to obtain candidate characters; screening target characters from all the candidate characters, and replacing the error characters with the target characters to obtain a target text. Through the character matching on the character strokes of each error character, the character matching using stroke information is realized, and the accuracy of text error correction in the information input process in the insurance field is improved.

The application provides an OCR text correction method and device. A correct probability of a Chinese character at each position in to-be-corrected text is obtained. A character shape structure word embedding vector of the Chinese character at each position in the to-be-corrected text is obtained. An input word embedding vector of the Chinese character at each position is calculated according to the word embedding vector corresponding to the Chinese character at each position in the to-be-corrected text, the correct probability and the character shape structure word embedding vector, and an input word embedding vector set is formed. The input word embedding vector set is input into a second encoder to form a coded input word embedding multi-dimensional vector. The coded input word embedding multi-dimensional vector is input into a correction layer to obtain corrected text. The correction layer is provided with a correction neural network formed by training of a Chinese corpus. The OCR text correction method and device provided by the application no longer selects a candidate word from an existing confusion set for correction, reduces the occurrence of missing words, and improves the accuracy of correction.

The present application relates to the technical field of computer information retrieval, and particularly relates to a digital publishing platform supplementary character retrieval method and system, the present application provides a kind of digital publishing platform supplementary character retrieval method and system, retrieve function is called out by user end, supplementary character feature is matched with supplementary character database, corresponding glyph is extracted according to matching data, after obtaining search term, the search term data uploaded to central server is processed by indexing word division data in supplementary character database and preset algorithm based on supplementary character, key word full-text retrieval is carried out by word division, and the search result is displayed in user end, solve the supplementary character of prior art and can only be expressed by screenshot form, picture is difficult to form effective index, cannot be directly added to full-text retrieval engine, leading to related digital publishing application cannot be based on supplementary character to realize key word retrieval, pinyin, initial letter retrieval and other problems.

The invention discloses a printed text content display method and device based on cross-platform consistency, and the method comprises the steps: obtaining a to-be-printed character string and a defined text object description protocol, and obtaining a text attribute structure according to the defined text object description protocol; obtaining the resolution ratio of the ink-jet printer nozzle, and enabling the resolution ratio of the PC upper computer to be consistent with the resolution ratio of the ink-jet printer nozzle according to the resolution ratio of the ink-jet printer nozzle; adopting a text shaping engine to obtain corresponding font list information according to the text attribute structure; adopting a font rasterization engine to obtain corresponding font contour information according to the font list information and the text attribute structure; obtaining the pixel width and height of the whole line of text according to the font list information and the font contour information; according to the obtained font list information, font contour information and resolution parameters, drawing to a display screen by adopting a Skia rendering engine; therefore, cross-platform measurement and drawing consistency of the text content of the ink-jet printer is realized.

The invention discloses a printing control method and electronic equipment. The method comprises the following steps: acquiring to-be-printed data; the coding type and the language type of the to-be-printed data are analyzed; selecting word stock processing channels based on the coding type and the language type, wherein the word stock processing channels comprise a dot matrix word stock processing channel and a vector word stock processing channel containing an integer rendering process; the to-be-printed data is converted into printing dot matrix data through the word stock processing channel; printing is completed based on the printing dot matrix data. According to the method, by establishing a dot matrix and vector mixed font library processing architecture and based on coding and language type intelligent routing, the problems of font deformation and stacking display of a complex language family are solved by utilizing a shaping rendering engine while high-speed printing of basic characters is guaranteed, and the balance of the printing speed and the output quality in a resource limited environment is realized.

The present disclosure relates to a font recognition method, device, readable medium and electronic equipment. The font recognition method divides the image to be recognized into a plurality of sub-images through the predetermined font recognition model, and obtains the first image corresponding to each sub-image. Image features, determine the second image features corresponding to the image to be recognized based on the first image features corresponding to each sub-image of the image to be recognized, the second image features include each sub-image of the image to be recognized and the context-related features of other sub-images are used to determine the font type corresponding to the target text based on the second image features. In this way, the image to be recognized can be described more comprehensively and accurately based on the correlation between each word image and other sub-images. This can effectively improve the accuracy of font recognition results and also effectively improve the font recognition rate.

This application discloses a method and apparatus for calligraphy and painting recognition based on deep learning, belonging to the field of artificial intelligence and digital image processing technology. The method includes: acquiring images of calligraphy and painting works; preprocessing and enhancing restoration by introducing a generative adversarial network with an edge-preserving loss function to preserve the texture of the ink marks; extracting the character structure features of the signature, inscription, and seal areas; inputting them into a specially trained optical character recognition and semantic analysis hybrid model; during decoding, weighted fusion of visual feature vectors and contextual semantic probability vectors to output initial recognition text; combining a calligraphy dictionary database, performing semantic error correction and matching through a joint calculation formula of quantified character topological edit distance and semantic probability to obtain and output the target recognition text. This solves the problems of failure due to lack of topological repair capability for damaged pixels and inability to hit the correct characters, achieving high-precision recognition and deep knowledge interpretation of calligraphy and painting text.

This application discloses a method, apparatus, electronic device, and storage medium for character stroke extraction. The method includes: acquiring the glyph curve of the character to be processed; performing discrete sampling on the glyph curve to obtain a discrete point set; selecting at least two stroke key points from the discrete point set to obtain a closed curve containing at least two stroke key points, thereby obtaining the character strokes. Through the above method, this application can improve the speed and accuracy of character stroke extraction.

The invention discloses a real ink mark driven facsimile authentic method based on submerged space diffusion, which comprises the following steps: firstly, collecting a target facsimile image and a large number of real ink mark sample images, carrying out preprocessing operation on the images, and constructing a standardized training set and a test set; secondly, extracting a contour diagram and a skeleton diagram from the target copy image, and mixing the contour diagram and the skeleton diagram in a linear mode to generate a continuous structural feature map; constructing an ink mark encoder, encoding the authentic mark sample, and extracting the pen and ink characteristics of the authentic ink mark sample; and finally, constructing a reinking network based on a latent space diffusion model, and finally converting the optimized data representation into a pixel image through a decoder to obtain a calligraphy image with natural ink marks and real strokes. The problems that in the prior art, copy images are inflexible in ink color and lack of natural ink color change are solved, meanwhile, stability of a font skeleton and refinement of the stroke of the to-be-restored area are guaranteed, and authentic transformation work is achieved.

Methods and systems are provided for automated reconstruction of glyph spacing properties. In embodiments described herein, a document comprising a set of glyphs of a text object is accessed. A tracking pattern is determined for the set of glyphs that identifies spacing of the set of glyphs in the document based on a position of a glyph of the set of glyphs in the document relative to a corresponding position of the glyph when automatic kerning is applied. The spacing of the text object is reconstructed in the document by applying the tracking pattern to the set of glyphs.

A kind of Chinese character font generation method, including constructing target font dataset, constructing source font dataset, constructing training dataset, constructing font style transfer network model, training font style transfer network model, Chinese character font automatic generation, the skeleton extraction module in the font style transfer network model of the present application, utilize font skeleton information to guide the generation of Chinese character font, make the Chinese character form generated more complete, the present application greatly optimizes the problems such as parameter explosion and training difficulty increase caused by attention mechanism superposition, improves the problems such as font stroke adhesion, detail loss, even lack of stroke generated by existing method, while reducing the time of training network, improve the quality of Chinese character font generation;The input font style transfer network model of the present application can also be other fonts such as handwritten font, provide convenience for the generation and diversification development of Chinese calligraphy.

The present application belongs to the field of computer vision and vector image generation, and particularly relates to a vectorized Chinese character pattern generation method based on a large model. The method comprises the following steps: step one, data preprocessing; step two, stroke feature extraction; step three, feature vector discretization coding; step four, stroke sequence combination; step five, feature coding prediction; step six, coding feature restoration; and step seven, stroke restoration. The present method combines Chinese character stroke order information, and on the basis of being able to complete Chinese character stroke generation to form a complete vector image, can also generate continuous semantic Chinese character vector text content in the form of continuously predicting the next character stroke, can complete missing character shapes in the case of incomplete font library, realize creative vectorized Chinese character text generation, make contributions to the fields of artistic character shape generation, Chinese character font auxiliary design, advertisement text design, etc., reduce labor costs, and realize efficient generation.

Systems or methods for presenting or generating an animated realization of one or more letter glyphs based on progressively exposing a continuous mapping of brush head shapes following a letterform stroke order. A machine-learning model is used to generate animation metadata from the one or more letter glyphs. The animation metadata including a centerline, the continuous mapping of brush head shapes that is used to sweep along the centerline, and the letterform stroke order to generate the animated realization of the one or more letter glyphs.

Embodiments of the present application provide a method and device for determining word representation vectors, electronic equipment and computer readable storage medium, and belong to the field of natural language processing. The method comprises: obtaining a set of word units of a text; obtaining a context vector of the text based on the set of word units; and predicting a next word of the text based on the context vector. The method for determining word representation vectors can effectively obtain a corresponding set of word units even for pictographic characters or languages evolved from pictographic characters that are prone to out-of-vocabulary words, thereby improving the accuracy of determining word representation vectors.

The invention discloses a deep learning-based harmonic generation method and terminal, and the method comprises the steps: obtaining a source language text of a to-be-converted language, and extracting a phoneme stream of the source language text; segmenting the phoneme stream according to a language rule of a target language to obtain a plurality of first syllables; obtaining a pronunciation library corresponding to the target language, matching a second syllable corresponding to each first syllable in the pronunciation library, and obtaining a font corresponding to the second syllable; establishing an association relationship between the source language text and all fonts, and training the initial model through the association relationship to obtain a target model; and when a to-be-converted text corresponding to the to-be-converted language is received, obtaining a homophonic font corresponding to the to-be-converted text through the target model. According to the method, the nonlinear mapping rule between the phoneme combination of the source language text and the target language pronunciation is obtained based on deep learning, so that when facing uncommon vocabularies and new words never appearing in a training set, the system still can generate accurate and reasonable homophonic phonetic notation through reasoning according to the learned general pronunciation logic.

The invention discloses a boundary perception and pre-training enhanced low-resource Tibetan language recognition method and system, and the method comprises the following steps: 1, carrying out the standardization processing of data and texts, and enabling the language knowledge and pre-training model enhanced low-resource Tibetan language voice to be converted into text. Unified modeling of Tibetan language boundaries, normal characters and morphemes is realized under a low-resource condition; remarkable and stable alignment is realized through a tsheg-perception adapter and three-head joint learning; language knowledge is fused into search in a learnable mode through the micro WFST soft constraint, and illegal font and language / case marking errors are greatly reduced. Training data are effectively expanded through teacher semi-supervised and coverage rate active learning, and cross-domain and accent robustness is improved; the end side maintenance cost is remarkably reduced through'adapter + atlas' hot updating; iCR and a boundary F1 are introduced as closed-loop indexes, so that the quality of an optimization target is consistent with that of a Tibetan Chinese character method, and the whole method has the advantages in engineering landing, interpretability and maintainability.

The application discloses a style visual text editing method, system, device and storage medium, which are corresponding solutions, and the related solutions aim to solve the style consistency problem existing in the image text editing of the existing diffusion model, extract style embedding information from the visual features of the glyph image and the input text image, and input the style embedding information as an enhanced style condition into the diffusion model, so as to realize fine control of the diffusion process, so that the diffusion model can generate a text image with high readability and style consistency, and can realize the keeping of the original text style or style migration based on a reference image.

The invention belongs to the field of petroleum logging, and discloses a logging data identification method and related equipment suitable for a logging software interface, and the method comprises the steps: collecting a to-be-identified logging software interface image, and carrying out the separation processing of the image with a symbol as a basic unit, thereby simplifying the identification process; and the horizontal projection data and the vertical projection data of each symbol image are calculated and compared with symbol characteristics learned in advance, so that the logging data can be quickly and accurately recognized, complex font sizes and font changes do not need to be considered, the recognition speed is remarkably increased, and the recognition efficiency is improved. The requirements of double requirements on rapid change and accuracy of data in a real-time logging process are met; by adopting the method, the problems that an existing logging data recognition mode is low in recognition speed and cannot adapt to rapid change of real-time logging data due to the fact that factors such as font sizes and fonts are considered can be solved, the recognition speed is high, and the recognition accuracy is high; and the accuracy and reliability of logging data acquisition are improved.

The present application relates to the field of thermal printer printing data processing and control, and particularly relates to a kind of vector printing thermal printer control method and system.The method comprises: obtaining vector font library file and printer parameter, and constructing font rendering session configuration structure;Parsing print task field, arranging character encoding queue, loading glyph outline data and executing prompt contour fine tuning, anti-diffusion prompt rule mapping and contour point cluster sampling, generating anti-diffusion prompt contour package structure;Contour scaling, rasterization gray coverage calculation and binarization processing are carried out, and point row bitmap data stream structure is obtained;Based on point row bitmap data stream, point row heating timing is arranged, and print head heating instruction sequence is generated.The present application effectively reduces the influence of thermal diffusion by anti-diffusion prompt rule mapping, improves print clarity, and improves paper feeding precision by using closed-loop control, enhances system reliability and traceability.

A visibility color recommendation system includes functionality operable to receive a user input defining a glyph to be displayed along with a digital image in a user interface. A color recommendation is generated, automatically and without user intervention, based on visibility of the glyph with respect to one or more colors of the digital image. The glyph is presented in the user interface as having a color from the digital image based on the color recommendation.

The application discloses a teaching self-media courseware generation method and system, which is applied to the technical field of intelligent education and comprises the following steps: obtaining first teaching characteristic information and second teaching structured information corresponding to an initial teaching outline; obtaining at least teaching images and teaching texts; obtaining image semantic analysis results of the teaching images based on at least first character shape information, second character pronunciation information and third collocation information of all roots; performing natural language analysis and processing on the teaching texts to obtain text semantic analysis results; sequentially performing semantic alignment and semantic correction processing on the image semantic analysis results based on the text semantic analysis results to obtain image semantic correction results; reconstructing the teaching images based on at least a semantic deviation degree to obtain teaching image resources; and integrating the teaching texts and the teaching image resources according to the second teaching structured information to obtain a self-media courseware. The method disclosed by the application can generate a high-quality courseware with complete content logic structure.

The present disclosure belongs to the field of artificial intelligence, and proposes a scene text image generation method, system, device and storage medium, comprising: obtaining input prompt text for scene text image generation, and generating glyph images and initial background images based on the input prompt text respectively; positioning the text position in the initial background image to generate a mask image; introducing a scene image, and transforming the glyph image based on the scene image, the initial background image and the mask image to obtain a deformed text image, wherein the scene image is determined according to the initial background image and the mask image; generating an auxiliary image according to the deformed text image and the scene image, and generating a scene text image based on the auxiliary image. The proposed glyph separation effectively improves the accuracy of generated text; the proposed scene perception improves the realism of generated text; in view of the problem of visual quality decline caused by directly fusing glyph and background images, the proposed text repair realizes multi-language scene text image generation.

This invention relates to an intelligent method and device for restoring damaged stone inscriptions. The method includes: constructing an ancient script character database based on images of intact inscriptions; calculating the similarity between the input damaged stone inscription and the glyph samples in the ancient script character database, using structural consistency, contour similarity, skeletal correspondence, and stylistic similarity as criteria, and calculating the weighted similarity of the above four criteria; outputting k glyphs from highest to lowest according to the weighted similarity of the above four criteria, denoted as Top-k candidate characters; and reordering the Top-k candidate characters based on the context of the inscription, the relationship between adjacent characters, and common character usage rules, and outputting them to provide researchers with auxiliary references for restoration. The intelligent restoration method for damaged stone inscriptions provided by this invention only requires a small number of high-quality inscription samples for training when constructing the ancient script character database, and models are built based on the stone inscription samples, which has the advantages of strong practicality and high restoration accuracy.

The embodiment of the invention provides a text rendering method, a font transformation method and related equipment, and belongs to the technical field of graphic rendering. The text rendering method comprises the following steps: receiving a text editing instruction through a Web drawing board, and identifying a target character and a font style in the text editing instruction; analyzing a vector font file corresponding to the font information, and extracting original contour vertex data of the target character; if the font information is inconsistent with the font information of the vector font file, transforming the original contour vertex data according to the font information, and storing the transformed contour vertex data into a buffer region corresponding to the color information; and calling a webpage graphics library drawing instruction, and carrying out combined drawing on the vertex data in the buffer area on the Web drawing board. Different font styles can be realized based on one font file.

According to the method, the codes with the combination of the front shape and the rear shape are formed by splitting pronunciation and fonts in multiple regions for character symbols, the unique symbols are quickly retrieved, sequential searching of human eyes is reduced, and the problems that in an input method scheme, homophones are too many, codes are too long, repeated codes are too many, and learning cost is too high are solved.