字幕增强中国图象图形学报.DOC

中图法分类号：TP391.4 文献标识码：A 文章编号： Logistic视频字幕增强模型 李钦瑞1，吕学强1，李卓1，刘坤2 1.北京信息科技大学网络文化与数字传播北京市重点实验室，北京 100101； 2.北京拓尔思信息技术股份有限公司，北京 100101 摘 要：目的：为提高复杂背景下的视频字幕在OCR中的识别率，需要对提取的视频字幕进行有效地字幕增强。该文首次将Logistic模型应用到视频字幕增强中，提出了基于Logistic模型的融合多帧信息的视频字幕增强方法。方法：对字幕进行检测与跟踪，将出现在连续多帧中的同一字幕片段进行对齐；通过分析字幕片段在多帧中信息，提出字幕背景在时域上的变化特征、背景和字幕文本的固有特征，并将三个特征进行量化与融合，构建适用于字幕增强的Logistic模型，实现对视频字幕的增强。结果：对含阴影或描边效果的特殊复杂背景字幕、普通复杂背景字幕、单一背景字幕分别进行实验，增强后的字幕在OCR软件中的识别正确率分别为81.76%、97.13%、98.19%，与对比方法比较均有一定的提高。结论：Logistic model for video caption enhancement Li Qinrui1, Lv Xueqiang1, Li Zhuo1, Liu Kun2 1.Beijing Key Laboratory of Internet Culture and Digital Dissemination Research, Beijing Information Science and Technology University, Beijing, 100101； 2. Beijing TRS Information Technology Co., Ltd，Beijing, 100101 Abstract: Objective: Video caption contains abundant information related to the video content. Recognizing text in images is the premise of making full use of this information. Although the OCR software recognition accuracy has been improved, the video caption with complex background couldn’t be recognized well yet. Therefore, in order to improve the recognition accuracy, the extracted caption shall be enhanced which can reduce the complexity of caption background and improve the contrast between background and text. This paper proposes a method of fusing multi-frame information to realize caption enhancement based on Logistic model. Method: Logistic curve is a common form of S-type curve which either end converge to a constant. By counting and analyzing distribution proportion of different pixel values in single background caption, we establish a proper Logistic model that the output can be used as enhanced caption’s pixel values and their distribution proportion shall be kept consistent with single background caption generally. According to the convergence of Logistic model, majority of pixel values can be assigned to 0 or 255, and a small quantity of gray points can be taken as transitions of b