30 results about "Stroke extraction" patented technology

The present invention discloses a method and a device for touching character segmentation in character recognition, and belongs to the character recognition field. The method comprises the steps of carrying out preprocessing to obtain connected domains, an average character width and an average character height of a row image; carrying out connected domain analysis, marking touching connected domains, carrying out stroke extraction for selected touching connected domains, carrying out segmentation point detection of the touching connected domains to obtain pre-segmentation points, and saving character blocks for non-touching connected domains; merging extracted strokes according to the pre-segmentation points to obtain the character blocks; saving the character blocks and turning to a next connected domain for carrying out touching determination, and outputting a character block sequence after all connected domains having been traversed; and merging the character blocks according to reference information and outputting recognition result. The method and the device provided in the invention merge the strokes according to the pre-segmentation points to obtain the character blocks, guaranteeing that segmentation points in a larger scope can be detected, and take the mode that contour information is used to predetect the segmentation points as a parameter in merging, avoiding merging mistakes caused by merging correct segmentation points.

The invention discloses an off-line Chinese character stroke extraction method based on template matching. The off-line Chinese character stroke extraction method comprises the following steps: stroke codes are established for all strokes of Chinese characters and serve as templates; a code is established for a given Chinese character according to the types and numbers of the strokes and serve as a template; the framework line of the given off-line Chinese character is extracted; the framework line is dismantled into four types of stroke segments according to the directions of four strokes namely transverse, vertical, left-falling and right-falling; the stroke segments, which can possibly form new strokes through end-to-end connection, are connected to form new quasi-strokes, and the new quasi-strokes are identified according to the directions and the stroke segment composition features; the quasi-strokes, the stroke code template of the given Chinese character and the Chinese character code templates are matched to confirm all strokes of the off-line Chinese character. With adoption of the method capable of extracting the strokes of the off-line Chinese character through the framework, confirmation of each stroke of the off-line Chinese character can be facilitated, and support can be provided for giving dynamic writing attributes to the strokes of the off-line Chinese character and even the whole off-line Chinese character and for researching and developing the generation technology of a new Chinese character font library.

The invention discloses a method for extracting automatically character strokes based on splitting and matching, and belongs to the technical field of pattern recognition and information processing. The method comprises the steps as follows: Step one, splitting a character into multiple stroke components according to a character structure; Step two, extracting a skeleton diagram of the stroke components, finding junctions in the skeleton diagram, then calculating PBOD curves of points which are in the neighborhoods of the junctions in a binary image, and extracting junction areas; Step three, extracting features for the stroke components, wherein, the features are matched with features of the stroke components in a standard library; and Step four, combining stroke segments according to the matched result and obtaining strokes of the character. According to the invention, the satisfactory extraction effect can be achieved for the characters with different fonts, the deficiencies that the character stroke extraction accuracy is not ideal and the character stroke extraction speed is slow can be improved effectively, and the stroke extraction accuracy is also high increasingly after the standard character library is complete increasingly.

The invention relates to an automatic extraction method for character strokes of vector fonts. The method includes steps of 1) performing skeleton extraction on to-be-extracted target fonts and template fonts corresponding to the target fonts and having stroke classification information in a template database and obtaining data point set and template point set; 2) taking the data point set, the template point set and the stroke attachment of the template point set as input, performing non-rigid point set registration based on structural information and obtaining the stroke attachment of the data point set; 3) converting the stroke attachment of the data point set to attachment of data outline segments so as to enable each outline segment forming a character shape outline to have corresponding attached strokes; 4) connecting the generated outline segment and obtaining a final vector stroke extraction result. According to the invention, font stroke extraction can be realized accurately and no manual intervention is needed. Automatic algorithm operation can be achieved and mass operation of vector stroke extraction can be facilitated.

The invention provides a stroke width extraction method and device and a character recognition method and system and solves the problem of accuracy in stroke extraction. The stroke width extraction method includes: extracting an original connected component of a stroke and allowing the component to correspond to a candidate stroke width; subjecting to the original connected component to corrosion calculation, calculating outline perimeter of the connected component before each time of corrosion, and forming a perimeter histogram; using the candidate stroke width that the connected component corresponds to after each time of corrosion calculation to perform difference calculation on the perimeter histogram so as to obtain stroke length of each candidate stroke width; forming a stroke width histogram by the stroke lengths that the candidate stroke widths correspond to; determining whether the original connected component is a character region or not according to a maximum of the stroke width histogram; if yes, determining the stroke width of the character region according to the maximum. The stroke width extraction method and device and the character recognition method and system have the advantages that stroke width extraction can be more accurate and calculation speed is increased.

The invention relates to a Chinese character font generation method based on a conditional generation antagonistic network, which relates to Chinese character font generation. Stroke information of Chinese character bitmap is extracted by using central skeleton extraction and coherent point drift method. The new style strokes are generated by the conditional antagonism generation network. A methodfor extract strokes by use Lian Zhouhui is broadened, and that stroke extraction is expanded from handwritten font without thick and thin information to Chinese character bitmap with thick and thin information; In font generation, the proposed method has the advantages of diverse generation results, simple training process, no special prior knowledge, and less manpower and time consumption.

The invention discloses a method for automatically extracting strokes of a calligraphy single character. The method comprises the steps of 1, performing graying and binarization processing on a calligraphy single-character picture, and obtaining a contour of a single character according to information of four neighborhood pixel points of black pixel points in the picture after binarization; 2, carrying out angular point detection on the contour obtained in the step 1, selecting a constraint edge set and a constraint point set, and carrying out constraint Delaunay triangulation; 3, calculatingthe minimum distance between any two connecting triangles containing more than two corner points in the step 2, and determining a cross region and stroke sections connected with the cross region; and4, carrying out continuity analysis on the stroke segments connected with all the intersection regions in the step 3, connecting the stroke segments meeting conditions to obtain a single stroke contour, carrying out contour filling, and finally completing extraction of strokes. According to the method, the intersection area can be accurately found in the stroke extraction of the calligraphy singlecharacter; and the composite strokes can be well extracted.

The invention discloses a soft-pen Chinese character conversion method based on the edition of handwriting characteristics. The above method can be used for extracting the handwriting characteristics of Chinese characters in an inscription/soft-pen copybook and generating soft-pen Chinese characters that are customized in handwriting characteristics. The method comprises the steps of firstly, selecting an appropriate noise-reduction algorithm, a binary operation and a contour smoothing method to pre-process an original inscription image and improve the contrast ratio of the inscription image relative to a background; secondly, positioning each single Chinese character through the segmenting and normalizing operation; thirdly, realizing a stroke extraction and optimization algorithm based on contour triangulation, constructing a triangular mesh for Chinese characters based on the above pre-processing operation, and marking out stroke segments; at the same time, dividing stroke sets based on handwriting characteristics; and finally, providing a Chinese character distortion generating algorithm based on stroke characteristics. According to the technical scheme of the invention, in combination of a customized characteristic matrix composed of stroke center lines, width, contraction speed, flexibility and other factors, and inputted distortion requirements, the characteristic matrix is modified. Therefore, Chinese character images of different handwriting characteristics can be generated.

The invention discloses a Chinese character Webfont generation method. The Chinese character Webfont generation method comprises the steps of submitting a character request character string S, obtaining font descriptors d corresponding to each Chinese character in the S by searching a DCDML dynamic description library file to obtain a font description set D; carrying out stroke extraction on the font descriptors d in sequence to obtain a point set P forming strokes; successively connecting the point coordinates of the P to form the strokes; integrally displaying the connecting information of all point strokes in the set P, and generating corresponding Webfont Chinese characters f of the font descriptors d; carrying out local optimization on the f; obtaining a Webfont character string corresponding to the character request character string S, and returning to a user. According to the Chinese character Webfont generation method disclosed by the invention, the font counting process is carried out on a server side, so that the requirement for the server side is relatively low, and the method is extremely suitable for being used on mobile terminals.

The invention provides an image recognition processing method and device, electronic equipment and a storage medium, which is applied to the map field. The method comprises the steps of recognizing at least one character area image from an image to be processed, conducting character recognition processing on each character area image, and obtaining character library characters corresponding to each character area image in a character library; performing stroke extraction processing on the character area image to obtain a stroke extraction result of at least one stroke frame in the character area image; performing stroke alignment processing based on the stroke extraction result of each stroke frame in the character area image and the character library characters corresponding to the character area image to obtain a stroke corresponding relation between strokes of each stroke frame in the character area image and strokes of the character library characters corresponding to the character area image; through the method and the device, the strokes can be accurately and efficiently recognized from the image comprising the characters and correspond to the strokes of the characters in the character library.

The invention belongs to the field of computer vision and pattern recognition, particularly relates to a Chinese character image stroke extraction method and system based on a full convolutional neural network, and aims to solve the problem that handwriting character strokes written freely are difficult to extract. The Chinese character image stroke extraction method comprises the following steps:carrying out region extraction on an acquired Chinese character image; carrying out skeletonization operation on the overlapped area and the non-overlapped area; calculating the coherence between anystroke segments of the overlapped area after skeletonization; and connecting the stroke segments belonging to the same stroke in the overlapped area, and combining the stroke segments directly connected in the non-overlapped area into a complete skeleton form stroke. On one hand, stroke extraction of handwritten Chinese characters can still be achieved under the condition that strokes of the handwritten Chinese characters which are written freely are overlapped; and on the other hand, a character synthesis method is adopted for obtaining a training sample, and different marking information ofthe training sample in different tasks is attached, so that the labor cost is greatly saved.

The invention discloses a MaskRCNN-based Chinese character stroke segmentation and extraction method. The method comprises the following steps: forming a stroke segmentation data set, analyzing fonts, and generating a training set and a verification set; constructing a MaskRCNN network model, wherein the MaskRCNN network model comprises a backbone network, a regional suggestion network and a branch network; presetting initial parameters and iteration times of a MaskRCNN network model, and performing image enhancement on the training set to supplement data; training a MaskRCNN network model, updating network parameters through the training set, and selecting and storing optimal parameters through the verification set; inputting a to-be-detected Chinese character image into the trained network model to obtain a stroke segmentation extraction result; according to the method, the Mask RCNN algorithm in the instance segmentation field is innovatively used in the stroke segmentation extraction task of the Chinese character, and compared with a stroke extraction method, the processing effect and efficiency are greatly improved.

The invention belongs to the technical field of image processing, particularly relates to a handwritten Chinese character homonymous stroke extraction method based on position constraint, which aims to solve the problem of low extraction applicability and accuracy due to the fact that stroke arrangement is not considered in extraction of homonymous strokes of handwritten Chinese characters. The method comprises the following steps of acquiring a handwritten Chinese character image and a reference Chinese character image, respectively centrally placing foreground areas of the handwritten Chinese character image and the reference Chinese character image in the two empty images to serve as a first image and a second image, obtaining a segmented connected region of the first image, traversing the segmented connected region of the first image, and adjusting the position of the reference stroke in the second image, obtaining a homonymous stroke connected region in the first image, and outputting a point set corresponding to the homonymous stroke connected region and the outer contour thereof. The method solves the problem that extraction of homonymous strokes of handwritten Chinese characters needs to pay attention to overall fonts and consider stroke arrangement, and improves applicability and accuracy of extraction of homonymous strokes of handwritten Chinese characters.

The invention provides a copybook generation system based on a deep neural network, and the system comprises a font preparation system, a stroke extraction system, a generation system, and a construction model evaluation system. The font preparation system transmits handwritten fonts to the system in the form of photos for a user. The stroke extraction system is used for respectively extracting single strokes or connected strokes of fonts in a picture, the generation system is used for learning according to the association of the single strokes and the connected strokes to form a font model, and the model evaluation system is constructed for evaluating the quality of the font model so as to form a closest font library. So the method has the advantage that corresponding full-text font library fonts can be formed according to different fonts of each person in a short time.

The invention relates to the field of information security, and discloses a method for extracting and authenticating stroke extraction and connection characteristics in online handwriting authentication. The method comprises the following steps: carrying out stroke extraction and pen feature pre-learning; obtaining the probability of occurrence of the false extraction and continuous extraction features in the crowd; registering stroke extraction and connection stroke characteristics, counting the occurrence frequency of stroke virtual extraction and connection stroke characteristics of a registered user, finding out stable virtual extraction and connection stroke according to the occurrence frequency, authenticating the stroke extraction and connection stroke characteristics, calculating the occurrence probability of stable and consistent combination of the virtual extraction stroke and the connection stroke in the crowd, and taking the probability as a basis for further judgment. Theinvention discloses a method for extracting and authenticating stroke extraction and connection characteristics in online handwriting authentication. According to the method, the characteristics of the virtual extraction pen or the continuous extraction pen in the local strokes can be effectively extracted, the occurrence probability of the stable and consistent combination of the virtual extraction pen and the continuous extraction pen in the crowd is estimated, a writing mode which violates writing standards and is stable can be recognized, and the handwriting authentication performance is improved.

The invention discloses a character stroke extraction method and apparatus, an electronic device and a storage medium. The method comprises the steps of obtaining a font curve of a to-be-processed character; performing discrete sampling on the font curve to obtain a discrete point set; at least two stroke key points are selected from the discrete point set, a closed curve at least containing the two stroke key points is obtained, and strokes of the character are obtained. In this way, the speed and accuracy of character stroke extraction can be improved.

The invention discloses a method for extracting strokes of handwritten Chinese characters. The method comprises the following steps of: preprocessing a character image; calculating direction point angle features of the pixels; calculating direction point angle features of the inflection point group; and stroke extraction based on the stroke skeleton and the contour. According to the method, off-line handwritten Chinese characters can be converted into corresponding strokes, then downstream tasks such as Chinese character handwriting recognition and writer identity recognition are carried out based on the strokes, and the interpretability of the whole processing flow can be greatly improved. By using the method provided by the invention, off-line handwritten Chinese characters can be more accurately converted into corresponding strokes, so that the accuracy of downstream tasks such as subsequent Chinese character handwriting recognition and writer identity recognition is improved.