3-D Reconstruction of Indoor Landscape Panorama Based on RGB-D Video

2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE) Pub Date : 2020-09-27 DOI:10.1109/ICISCAE51034.2020.9236888

Cai Chen

{"title":"3-D Reconstruction of Indoor Landscape Panorama Based on RGB-D Video","authors":"Cai Chen","doi":"10.1109/ICISCAE51034.2020.9236888","DOIUrl":null,"url":null,"abstract":"In order to improve the analysis and recognition ability of indoor landscape full-motion video images, super-resolution reconstruction of indoor landscape full-motion video images is needed. A virtual reality reconstruction method of indoor landscape full-motion video images based on RGB-D video analysis is proposed. The color feature decomposition rules of RGB images are adopted to decompose multi-pixel features of indoor landscape full-motion video images, information fusion is carried out according to the edge contour feature distribution of the images, multi-dimensional pixel spatial distribution feature quantities of the indoor landscape full-motion video images are extracted, super-resolution feature reorganization and block region matching of the indoor landscape full-motion video are carried out through a block grid region feature matching method, and an edge contour detection model of the indoor landscape full-motion video images is established. RGB decomposition of indoor landscape full-motion video image is carried out by wavelet multi-level structure decomposition method, pixel fusion model of indoor landscape full-motion video image is established, and super-resolution visual reconstruction of indoor landscape full-motion video image is realized by pixel region reconstruction method. The simulation results show that the super-resolution reconstruction of indoor landscape full-motion video images using this method has better visual expression ability and higher output peak signal-to-noise ratio.","PeriodicalId":355473,"journal":{"name":"2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICISCAE51034.2020.9236888","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

In order to improve the analysis and recognition ability of indoor landscape full-motion video images, super-resolution reconstruction of indoor landscape full-motion video images is needed. A virtual reality reconstruction method of indoor landscape full-motion video images based on RGB-D video analysis is proposed. The color feature decomposition rules of RGB images are adopted to decompose multi-pixel features of indoor landscape full-motion video images, information fusion is carried out according to the edge contour feature distribution of the images, multi-dimensional pixel spatial distribution feature quantities of the indoor landscape full-motion video images are extracted, super-resolution feature reorganization and block region matching of the indoor landscape full-motion video are carried out through a block grid region feature matching method, and an edge contour detection model of the indoor landscape full-motion video images is established. RGB decomposition of indoor landscape full-motion video image is carried out by wavelet multi-level structure decomposition method, pixel fusion model of indoor landscape full-motion video image is established, and super-resolution visual reconstruction of indoor landscape full-motion video image is realized by pixel region reconstruction method. The simulation results show that the super-resolution reconstruction of indoor landscape full-motion video images using this method has better visual expression ability and higher output peak signal-to-noise ratio.

查看原文本刊更多论文

基于RGB-D视频的室内景观全景三维重建

为了提高室内景观全动态视频图像的分析和识别能力，需要对室内景观全动态视频图像进行超分辨率重建。提出了一种基于RGB-D视频分析的室内景观全动态视频图像的虚拟现实重建方法。采用RGB图像的颜色特征分解规则对室内景观全动态视频图像进行多像素特征分解，根据图像的边缘轮廓特征分布进行信息融合，提取室内景观全动态视频图像的多维像素空间分布特征量;通过分块网格区域特征匹配方法对室内景观全运动视频进行超分辨率特征重组和分块区域匹配，建立了室内景观全运动视频图像的边缘轮廓检测模型。采用小波多层次结构分解方法对室内景观全运动视频图像进行RGB分解，建立室内景观全运动视频图像的像素融合模型，采用像素区域重构方法实现室内景观全运动视频图像的超分辨率视觉重构。仿真结果表明，利用该方法对室内景观全动态视频图像进行超分辨率重建，具有较好的视觉表达能力和较高的输出峰值信噪比。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2020 IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE)

自引率

0.00%

发文量