视频有效特征的提取是视频检索系统中的重要组成部分,它对于视频的检索、分析以及多媒体行业的发展具有积极的意义。本文以中外视频有效特征提取技术的发展现状为背景,简单阐述了本文的研究目的和意义,以及发展过程中存在的技术瓶颈,对视频有效特征的提取进行了系统的阐述。 本文首先从视频的有效特征入手,对视频的图像特征、文本特征和音频特征进行了简单的阐述。接着,着重的介绍了基于图像特征的视频提取,图像特征主要包括颜色特征、纹理特征、文字识别、运动特征和其他特征。在这章本毕业论文重点阐述了图像直方图的特征检索,对颜色特征直方图、颜色累加直方图、颜色局部累加直方图和改进的颜色直方图进行了详细的研究,并使用MATLAB等软件对图片进行了相关图像检索。与此同时,文本也介绍了纹理、运动等特征以及基于他们的视频特征提取技术。其次,本文也介绍了基于文本特征的视频提取,主要有文本区域定位和视频文本识别。最后,对本文进行了总结与展望,为今后的视频有效特征提取技术的发展提供一些参考。8340
关键词 视频特征 颜色特征 颜色直方图 文本特征
Title Extraction and experimental research about effective
visual feature of video
Abstract
Extraction and experimental research about effective visual feature of
video is an important part in the retrieval system of video. It has positive
significance for retrieval and analysis of video and multimedia industry.
We have a simple explanation about the image features, text features and
audio features of video. Then we focus on the video extraction based on
image features. The image features include color features, texture
features, character recognition, movement characteristics and other
features. In this chapter, we described the characteristics of the image
histogram retrieval, and we carry out a detailed study on the color
characteristics histogram, color cumulative histogram, the color of the
local cumulative histogram and color histogram improved. Also we have a
image retrieval on a picture using MATLAB. Finally, we do a summary and
outlook to provide reference for the development of the extraction of video
feature.
Keywords video features color features color histogram text features
目 次