首页> 外国专利> METHOD AND SYSTEM FOR SOUND SOURCE LOCALIZATION USING IMAGE INFORMATION AND STORAGE MEDIUM STORING PROGRAM TO REALIZE THE METHOD

METHOD AND SYSTEM FOR SOUND SOURCE LOCALIZATION USING IMAGE INFORMATION AND STORAGE MEDIUM STORING PROGRAM TO REALIZE THE METHOD

机译:利用图像信息和存储介质存储程序实现声源定位的方法和系统

摘要

PROBLEM TO BE SOLVED: To provide a method and a system for sound source localization where an image with monaural sound is automatically converted into an image with stereophonic sound and no sense of incongruity is caused in positional relation between the image and the sound source in a panorama image or the like. ;SOLUTION: In this sound source localization method, a division section 20 divides a received image with sound information into the sound information and the image. An image analysis section 40 uses information in an image intelligence database 30 to analyze an object in the image, a motion (position) of the object, and camera motion or the like from the image divided from the original image and to acquire image information. A sound source separation section 60 separates a sound source considered to be emitted from the object from the divided sound information on the basis of the acquired image information and information of a sound intelligence database section 50. A sound source localization section 70 relocates the separated sound source into a sound space proper to the video image by taking the acquired image information and a video display method such as display of a panorama image or the like in a reproduction section 90 into account. A synthesis section 80 synthesizes the relocated sound source and the image and the reproduction section 90 displays and reproduces the synthesized image with sound information by processing video display method.;COPYRIGHT: (C)2000,JPO
机译:解决的问题:提供一种用于声源定位的方法和系统,其中具有单声道声音的图像自动转换为具有立体声声音的图像,并且在图像和声源之间的位置关系中不会引起不协调感。全景图像等。 ;解决方案:在这种声源定位方法中,划分部分20将接收到的具有声音信息的图像划分为声音信息和图像。图像分析部40使用图像智能数据库30中的信息来从从原始图像划分的图像中分析图像中的对象,对象的运动(位置)以及照相机运动等,并获取图像信息。声源分离部分60基于获取的图像信息和声音情报数据库部分50的信息,从分割的声音信息中分离被认为是从物体发出的声源。声源定位部分70重新定位分离的声音。通过考虑获取的图像信息和诸如再现部分90中的全景图像的显示之类的视频显示方法,将声音信号输入到适合于视频图像的声音空间中。合成部分80将重新定位的声源和图像合成,并且再现部分90通过处理视频显示方法来显示和再现带有声音信息的合成图像。版权所有:(C)2000,JPO

著录项

  • 公开/公告号JP2000295700A

    专利类型

  • 公开/公告日2000-10-20

    原文格式PDF

  • 申请/专利权人 NIPPON TELEGR & TELEPH CORP NTT;

    申请/专利号JP19990095702

  • 发明设计人 MIYAGAWA KAZU;KOJIMA HARUHIKO;

    申请日1999-04-02

  • 分类号H04S7/00;H04N5/91;H04N7/18;

  • 国家 JP

  • 入库时间 2022-08-22 02:02:53

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号