A method and device for extracting text from an image

Abstract

The present invention relates to the technical field of picture processing, and specifically discloses a method and device for extracting pictures and texts in an image. The method includes: drawing a first picture with three channels according to a preset first picture drawing strategy; drawing a first picture with three channels according to a preset second picture The drawing strategy draws the second picture of three channels; draws the third picture of four channels according to the preset third picture drawing strategy; corresponds the pixels of the first picture, the second picture and the third picture one by one, and compares the pixels of the third picture Each pixel of the image is reset to generate the target image. The invention realizes the extraction of color pictures and texts, generates a target picture with translucency, further improves the anti-aliasing effect of the target picture, and can have a very good display effect when the target picture is superimposed on other pictures or videos; another On the one hand, the present invention is not limited to the extraction of patterns or characters, nor is it limited to the color effect of graphics and texts, and completely reforms the current situation that the prior art can only extract single color graphics and texts.

Description

technical field [0001] The invention relates to the technical field of picture processing, in particular to a method and device for extracting pictures and texts from an image. Background technique [0002] At this stage, the method of directly integrating the picture content with the video or another picture needs to separate the picture content and the picture background from the picture. The picture content refers to the picture and text in the picture, and the picture background refers to the content other than the picture content. . [0003] In the prior art, the image extraction method of the graphic engine is used to separate the image content and the image background, and a new image with the image content is generated from the image content. However, after the existing graphic engine is separated, new pictures with picture content will have poor anti-aliasing effect, and the display accuracy of new pictures with picture content is not high, which will obviously aff...

AI Technical Summary

Problems solved by technology

However, after the existing graphic engine is separated, new pictures with picture content will have poor anti-aliasing effect, and the display accuracy of new pictures with picture content is not high, which will obviously affect the use of new pictures with picture content effects, and the image extraction methods of the existing graphics and text engines generally cannot support non-standard artistic words and colored words (including gradient colored words and colored words with multiple colors, in which, there can be either a colored word in the picture content There can also be multiple colored characters), so that it is not easy to extract or separate the non-standard artistic characters and colored characters, such as the background color is white, the edge of the font in the foreground is black, and the middle is also white , the extracted foreground may appear fully transparent in the middle

Images