Unlock instant, AI-driven research and patent intelligence for your innovation.

Image Recognition Method for Latin Transliteration of Sanskrit Devanagari Printed Characters

A technology of image recognition and Devanagari, which is applied in character and pattern recognition, instruments, calculations, etc., can solve the problems of difficult character recognition and few researches on image recognition of Sanskrit characters, achieve high accuracy, high accuracy, and improve processing efficiency Effect

Active Publication Date: 2020-04-21
杭州佛学院
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, Sanskrit characters are difficult to recognize due to the complexity of their text structure, so there are relatively few researches on image recognition of Sanskrit characters.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image Recognition Method for Latin Transliteration of Sanskrit Devanagari Printed Characters
  • Image Recognition Method for Latin Transliteration of Sanskrit Devanagari Printed Characters
  • Image Recognition Method for Latin Transliteration of Sanskrit Devanagari Printed Characters

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0051] Such as figure 2 The Sanskrit Devanagari printed character picture shown above is divided into 6 Sanskrit Devanagari printed character blocks by the above-mentioned Sanskrit Devanagari printed character block segmentation method, encoded by the Jiugongge method and Tianzige method, and the pixel value and normalization processing are obtained , respectively calculate the nine-dimensional feature vector of each character block, as shown in Table 1 below:

[0052] Table 1

[0053]

[0054] Among them, the character vector numbering method adopts figure 1 The numbering method in is used for labeling; Sanskrit Devanagari printed character blocks 1~6 and connector 7 are respectively as follows figure 2 Shown; Wherein the Sanskrit Devanagari printed character blocks 1 to 5 adopt the Jiugongge method, and the Sanskrit Devanagari printed character block 6 adopts the matt method. The numbering method of the character vector x is: Jiugongge is from left to right, from bot...

Embodiment 2

[0057] Such as image 3 The shown Sanskrit printed character picture, according to the above method, obtains 11 character blocks by character block segmentation (according to figure 2 The labeling method starts from the bottom of the connector, from left to right are Sanskrit Devanagari printed character blocks 1~8, and then starts above the connector, from left to right is Sanskrit Devanagari printed character blocks 9~11), through the Jiugongge Or matts calculation method obtains the nine-dimensional feature vector of each character block, as shown in table 2 below:

[0058] Table 2

[0059]

[0060]

[0061] The Latin character recognition conversion after recognition is obtained by minimum variance matching: The recognition accuracy rate is 100%.

Embodiment 3

[0063] Such as Figure 4 Shown Sanskrit printed character picture, obtains 14 character blocks (numbering is the same as embodiment 2) by character block segmentation, obtains the nine-dimensional feature vector of each character block by Jiugongge and matts calculation method, as shown in table 3 below:

[0064] table 3

[0065]

[0066] The Latin character recognition conversion after recognition is obtained by minimum variance matching: karhicit, the recognition accuracy rate is 100%.

[0067] From the verification results of Examples 1 to 3, it can be known that the conversion method of the present invention has high precision and a correct rate of 100%, and has strong practicability.

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a Latin transliteration image recognition method for Sanskrit Devanagari printed characters. block segmentation to obtain some Sanskrit Devanagari printed character blocks; (2) identify the Sanskrit Devanagari printed character blocks obtained, and obtain the corresponding feature vectors of the Sanskrit Devanagari printed character blocks; (3) compare the obtained feature vectors with the standard Latin The feature vectors of the characters are compared, and according to the comparison results, the recognized Sanskrit Devanagari printed character blocks are converted into Latin characters. The invention realizes the direct conversion from the Sanskrit Devanagari noiseless printed character image to the corresponding Latin character, and has high accuracy, basically reaching 100% accuracy. The image recognition algorithm used in the invention is convenient and easy to implement, and has high efficiency. The technical solution provided by the invention is easy to implement and is suitable for practical document research applications.

Description

technical field [0001] The invention belongs to the field of computer character image recognition, and in particular relates to a method for recognizing Sanskrit Devanagari noiseless printed characters in Latin transliteration. Background technique [0002] Sanskrit is not only a beautiful language, but also has a detailed and complete conjugation system, which has aroused great interest of modern scientists. It is precisely because Sanskrit has perfect grammatical rules that it is very suitable for processing by computers. Utilizing the convenience brought by the current computer technology can promote our study and utilization of Sanskrit literature and accelerate the comparative study of Buddhist scriptures between Sanskrit and Chinese. As the first step in the research on computer recognition of Sanskrit Buddhist scriptures, it is necessary to realize the digitization and automatic recognition of Buddhist texts, and solve the key problem of low manual input efficiency i...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/34G06K9/46G06K9/62
CPCG06V30/158G06V10/267G06V10/44G06V30/293G06F18/22
Inventor 刘松柏
Owner 杭州佛学院