Detection method and detection device for character region

Abstract

The invention discloses a detection method and a detection device for a character region. The detection method for the character region comprises the following steps: carrying out corner point detection on a target image so as to obtain corner point high confidence coefficient regions of the image; carrying out edge detection on the target image so as to obtain an edge communication region of the image; judging whether any detected corner point high confidence coefficient region meets preset regular shape conditions; if yes, determining the region to serve as the character region; if not, further judging whether the coincidence proportion of the region to the edge communication region reaches a preset threshold value; if yes, determining the region to serve as the character region. With the technical scheme, by adopting a manner that the corner point detection is combined with the edge detection, the character region on the image is determined; under the conditions of complicated character background and irregular character regions, the character region can be detected more effectively.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and a device for detecting a character area. Background technique [0002] The development of multimedia technology has made the forms of expressing information on computers more abundant, which has changed the application fields of computers to a large extent. Among them, images (Images) are widely used media, especially with the development of image processing technology, images have become an important form of information bearing in the field of computers and the Internet. [0003] Compared with text, images have a stronger ability to express, but relatively, for computers, the work of understanding an image has become more complicated. However, if the image contains text content, these texts can often provide very rich information for the understanding of image semantics, so extracting text from images correctly can become a shortcut to understand image sem...

AI Technical Summary

Problems solved by technology

[0004] However, in practical applications, many images are noisy. For example, some photos taken by non-professional users are often seriously affected by light and prone to blurring. There are also some web images. Although the image itself is of good quality, the background of the text is very complicated. , the text area is often irregular, which leads to the serious interference of the detection method based on connected domain analysis, and the extracted connected domain often appears broken or the background sticks, which seriously affects the accuracy of text detection; text detection through machine learning method It is often limited by the training samples and cannot be extended beyond the samples. Moreover, it is currently difficult to be practically applied to the text type of such a large font library as Chinese characters, both in terms of detection speed and implementation cost.

Images