Research Article
A Parallel Drone Image Mosaic Method Based on Apache Spark
@INPROCEEDINGS{10.1007/978-3-030-48513-9_25, author={Yirui Wu and Lanbo Ge and Yuchi Luo and Deqiang Teng and Jun Feng}, title={A Parallel Drone Image Mosaic Method Based on Apache Spark}, proceedings={Cloud Computing, Smart Grid and Innovative Frontiers in Telecommunications. 9th EAI International Conference, CloudComp 2019, and 4th EAI International Conference, SmartGIFT 2019, Beijing, China, December 4-5, 2019, and December 21-22, 2019}, proceedings_a={CLOUDCOMP}, year={2020}, month={6}, keywords={Distributed mosaic method Parallel processing Apache Spark Big data Drone aerial image}, doi={10.1007/978-3-030-48513-9_25} }- Yirui Wu
Lanbo Ge
Yuchi Luo
Deqiang Teng
Jun Feng
Year: 2020
A Parallel Drone Image Mosaic Method Based on Apache Spark
CLOUDCOMP
Springer
DOI: 10.1007/978-3-030-48513-9_25
Abstract
MapReduce has been widely used to process large-scale data in the past decade. Among the quantity of such cloud computing applications, we pay special attention to distributed mosaic methods based on numerous drone images, which suffers from costly processing time. In this paper, a novel computing framework called Apache Spark is introduced to pursue instant responses for the quantity of drone image mosaic requests. To assure high performance of Spark-based algorithms in a complex cloud computing environment, we specially design a distributed and parallel drone image mosaic method. By modifying to be fit for fast and parallel running, all steps of the proposed mosaic method can be executed in an efficient and parallel manner. We implement the proposed method on Apache Spark platform and apply it to a few self-collected datasets. Experiments indicate that our Spark-based parallel algorithm is of great efficiency and is robust to process low-quality drone aerial images.