Dense Computing Task Analysis of Multi-view Matching Method and GPU Implementation

被引：0

作者：

Dai, Chenguang ^{[1
]}

Ji, Song ^{[1
]}

Yang, Jingyu ^{[2
]}

机构：

[1] Wuhan Univ, State Key Lab Informat Engn Surveying Mapping & R, Wuhan 430072, Peoples R China

[2] Key Lab Mapping Space Natl Adm Surveying, Mapping & Geoinformat, Beijing, Peoples R China

来源：

PROCEEDINGS OF INTERNATIONAL CONFERENCE ON COMPUTER VISION IN REMOTE SENSING | 2012年

基金：

中国博士后科学基金; 美国国家科学基金会;

关键词：

Multi-view Matching; Task; GPU; Parallel; Acceleration Rate;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

For the parallel processing of multi-view matching method of high resolution digital imagery, dense computing tasks and crucial parallel computing aspects were analyzed. And then, based on GPU implementation scheme of single stereo matching method, an improved GPU implementation scheme was proposed. Based on practical image datasets, experiments were done and the experimental results proved that the acceleration rate of the improved GPU implementation scheme could reach 20 to 100 for dense computing tasks, and compared with single CPU processing, possessed greater processing superiority.

引用

页码：183 / 188

页数：6