Method and device for recognizing characters from image

Abstract

The invention discloses a method and a device for identifying characters from an image. The method comprises the steps of obtaining a to-be-identified image set; respectively extracting a text regionfrom each image in the image set; extracting characters from each text area; and identifying the extracted characters to obtain a character identification result corresponding to each image. The technical scheme has the beneficial effects that the text area is extracted from the image instead of overall identification of the image, recognition accuracy can be greatly improved, the method is advantaged in that interference of similar character lines in the image is avoided, the identification effect in the image with relatively good fusion of posters and other characters and backgrounds is especially good, compatibility with multiple business scenes is realized, image data production, verification and operation cost is reduced, automatic exploration of content clues contained in the image is realized, and the method is of great help for image-based data mining.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to a method and device for recognizing characters from images. Background technique [0002] With the advent of the high-tech information age and the development of Internet computer technology, a large number of multimedia information has appeared in people's daily life in an exponential growth manner, and the sharp increase in image information has attracted more and more attention. However, Since the image stores the original information such as the color and brightness of the object in the form of pixels, and lacks a high-level description of the image content, it is difficult to automatically identify the content through computer analysis, understanding, retrieval and reuse of the image. [0003] At present, text recognition by means of OCR (Optical Character Recognition, Optical Character Recognition) has made great progress in technology, but there are still some de...

AI Technical Summary

Problems solved by technology

[0002] With the advent of the high-tech information age and the development of Internet computer technology, a large number of multimedia information has appeared in people's daily life in an exponential growth manner, and the sharp increase in image information has attracted more and more attention. However, Since the image stores the original information such as the color and brightness of the object in the form of pixels, it lacks a high-level description of the image content, and it is difficult to automatically identify the content through computer analysis, understanding, retrieval and reuse of the image.

[0003] At present, the use of OCR (Optical Character Recognition, Optical Character Recognition) to recognize text has made great progress in technology, but there are still some shortcomings in the direct recognition of images.

For example, images such as posters often contain text, and text detection is affected by factors such as language, text resolution, text spacing, distribution, text background, lighting, and color. Certain patterns and textures in such images It is difficult to distinguish the background from the text. It is difficult to directly recognize the image as a whole, and the recognition effect is not satisfactory.

Images