Method and device for segmenting characters from video image

Abstract

The invention discloses a method for segmenting characters form a video image, which comprises the following steps: positioning an original character image in the video image; extracting character stroke information from the original character image, obtaining a binary stroke image of the original character image according to the character stroke information; establishing a color model aiming at a character region in the binary stroke image, extracting a color layer image of the character region according to the color model; and removing excess backgrounds and noise by using an improved connection body analysis method with stroke masks so as to obtain the target character image. The invention has little computation and good property and can rapidly and accurately segment the characters from the video image of the complex backgrounds without machine learning.

Description

technical field [0001] The present invention relates to the field of video character recognition, in particular to a method for segmenting characters from video images and a device for segmenting characters from video images. Background technique [0002] As a high-level semantic feature, characters in video images play an important role in the understanding and indexing of video content. However, due to the low resolution of the video image, the complex background corresponding to the characters in the video image, and the large interference; and the traditional OCR system can only accept the binarized text image. It is one of the important research topics in the field of video character recognition. [0003] In the prior art, methods for segmenting characters in a video mainly include the following three categories: [0004] The first category, threshold-based methods: [0005] For details, please refer to the literature published by V.Wu, R.Manmatha, E.M.Riseman. 1999 ...

AI Technical Summary

Problems solved by technology

This type of method selects the point on the border of the text box as a seed, and uses the seed growth method to remove the connected body connected to the border. This method can achieve better results for images with better quality, but when the contrast is very low, or the text If it is the same color as the background and the border of the outer edge of the text is not closed, the stroke area in the text color layer image will be glued to the background area. In the connected domain analysis, the seed filling algorithm will erode the text

Therefore, this connected body analysis method is still not suitable for complex background, low resolution, and video images disturbed by noise

Images