A cloudbased remote sensing data production system. Remote sensing data parallel processing base on cloud platform. Parallel relative radiometric normalisation for remote. Apr 07, 2016 with the development of synthetic aperture radar sar technologies in recent years, the huge amount of remote sensing data brings challenges for realtime imaging processing. When scaling to largescale multicore computing, efficient parallelization of an esm, which demands fast parallel computing for longterm integration or climate simulation, becomes extremely challenging because of timeconsuming internal big data communication. The journal of parallel and distributed computing jpdc is directed to researchers, scientists, engineers, educators, managers, programmers, and users of computers who have particular interests in parallel processing andor distributed computing. Several studies have attempted to apply parallel computing to improve image mosaicking algorithms and to speed up calculation process. Parallel computing implementation for scansar mode data. Parallel kmeans clustering of remote sensing images based on mapreduce 163 kmeans, however, is considerable, and the execution is timeconsuming and memoryconsuming especially when both the size of input images and the number of expected classifications are large. Towards building a dataintensive index for big data.
Kaufmann, cluster analysis of moms02d2 data using spectral bands in combination with texture image, in proc. Big data parallel computing data intensive computing remote sensing data processing abstract with the recent advances in remote sensing rs techniques, continuous earth observation is generating tremendous volume of rs data. Current trends and future perspectives ieee conference publication. Remote sensing applications read specialized file formats that contain sensor image data, georeferencing information, and sensor metadata. Moderate resolution atmospheric transmission modtran is a commercial remote sensing rs software package that has been widely used to simulate radiative transfer of electromagnetic radiation through the earths atmosphere and the radiation observed by a remote sensor.
A straightforward approach is to divide the given remote sensing images into several partitions by a particular data partitioning method. According to the current level of artificial intelligence, it is impossible to establish a fully automatic system for parallel processing of remote sensing rs imagery. For this reason, the parallel computing rendering in remote sensing image processing is essentially necessary. In this paper, a fast method is presented for moisacing the remote sensing images from the uav without ground control points. The rapid development of generalpurpose graphic process unit gpgpu computing technology has. A case study of remote sensing data processing service by xicheng tan 1, liping di 2, meixia deng 2, jing fu 3, guiwei shao 3, meng gao 4, ziheng sun 2, xinyue ye 5, zongyao sha 4 and baoxuan jin 6. Building an elastic parallel ogc web processing service on. Research on the parallelization of the dbscan clustering. They introduce a generic model for parallel processing of remote sensing data, and, using a massive dataset of remote sensing data, they demonstrate the performance gains inherent in using spark.
Parallel and distributed dimensionality reduction of. Current tools for processing synthetic aperture radar sar data are both computation and memory intensive. However, for pcs, the limitation of hardware resources and the tolerance of time consuming present a bottleneck in processing a large. Because the texture memory is readonly memory, processed remote sensing image data still cannot be stored in the texture memory. Dec 19, 20 this paper describes software technologies for processing of earth remote sensing ers data received from latest generation satellites both foreign and domestic.
Parallel processing implementations of a contextual. Fortunately, the distributed parallel computing has provided a turning point to the quick calculation of remote sensing image. Remote sensing applications which may be addressed by neural networks using parallel processing technology. Heterogeneous networks of computers have rapidly become a very promising commodity computing solution, expected to play a major role in the design of high heterogeneous parallel computing in remote sensing applications. Moreover, operational remote sensing applications are not always fully automatic, but may require frequent interactions with the user, which requires a fast delivery of intermediate remote sensing results. A parallel processing algorithm for remote sensing classification. However, remotesensing rs algorithms based on the programming model are trapped in dense disk io operations and unconstrained network communication, and thus inappropriate for timely processing and analyzing massive, heterogeneous rs data. The application of high performance computing hpc technology to remote sensing data processing is one solution to meet the requirements of remote sensing real or nearrealtime processing capabilities. Remote sensing applications it is a software application that processes remote sensing data. In addition, these processes are performed concurrently in a distributed and parallel manner. Fpga, and so on technologies meet parallel computing needs for onboard image processing particularly. Recently, cloud computing has become a standard for distributed computing due to its advanced capabilities for internetscale computing, serviceoriented computing, and highperformance computing 14. Performance computing technology for remote sensing data processing and analyzing. Building an elastic parallel ogc web processing service on a cloudbased cluster.
Separating manual operation from remote sensing image. Inmemory parallel processing of massive remotely sensed data. Library of congress cataloginginpublication data gebali, fayez. High performance computing for satellite image processing and. Software technologies for processing of earth remote sensing. High performance computing in remote sensing is one of the first volumes to explore stateoftheart hpc techniques in the context of remote sensing problems. With remote sensing technology and the level of the continuous development, its data processing systems are becoming increasingly complex, management and maintenance costs. A parallel computing system of remote sensing images based. Cloud computing in remote sensing 1st edition lizhe wang. International geoscience and remote sensing symposium.
Apr 14, 2015 interest in image mosaicking has been spurred by a wide variety of research and management needs. Parallel computing rendering in specific remote sensing. Taking radar remote sensing into operational use requires processing of a relatively large number of large images within a limited time frame. High performance computing in remote sensing 1st edition. Cloud computing in remote sensing crc press book this book provides the users with quick and easy data acquisition, processing, storage and product generation services. This paper describes software technologies for processing of earth remote sensing ers data received from latest generation satellites both foreign and domestic. Uav remote sensing image processing and classification in.
These services use saas conception to provide an uniform manmachine interface for users data uploading and. Parallel processing of massive remote sensing images in a gpu architecture 209 storing remote sensing image results data. Most remote sensing algorithms have inherent parallelism. One of the most striking improvements in earth imaging satellite systems is the improved spatial resolution. It also develops a series of remote sensing data management and processing standards. Pdf remote sensing data parallel processing base on cloud. Performance optimization and evaluation for parallel. Research and performance analysis on parallel computing of. Heterogeneous parallel computing in remote sensing. Remote sensing images have been widely used in various fields, how to process remote sensing images in realtime has become an important problem. Pdf cloud computing in remote sensing researchgate. Based on global subdivision model gsm, this paper presented a framework of a parallel computing system of remote sensing images gspcs. Parallel processing of massive remote sensing images in a gpu architecture 199 using gpu to accelerate processing for all kinds of remote sensing, image processing algorithms have resulted in greater related research achievements. However, remote sensing rs algorithms based on the programming model are trapped in dense disk io operations and unconstrained network communication, and thus inappropriate for timely processing and analyzing massive, heterogeneous rs data.
Kmeans, however, is considerable, and the execution is timeconsuming and memoryconsuming especially when both the size of input images and the number of expected classifications are large. This special issue on big data in earth observation. Parallel computing on heterogeneous networks download ebook. In such technologies, developed algorithms run on a single node only but with multiple cores. This material is presented to ensure timely dissemination of scholarly and technical work. Pasm is a proposed multimicroprocessor for image processing. Applications of parallel processing linkedin slideshare. Wiley series on parallel and distributed computing.
Antonio plaza hyperspectral imaging parallel computing. The remote sensing data processing system is a typical distributed computing system, which concerns the parallel computing, spatial database, image processing and a series of computer technology. This paper develops several highly innovative heterogeneous parallel algorithms for information extraction from high. The german aerospace center applied the technology to remote sensing data processing. However, when the number of node increases, the consuming time decreases, and the efficiency will decrease at the same time. Pdf a parallel computing paradigm for pansharpening. Based on virtualization, computing resources and software can be.
With the continuous improvement of earth observation satellite sensors and computer techniques, satellite remote sensing data has exploded in recent years, and a. The concept of parallel computing is based on dividing a large problem into smaller ones and each of them is carried out by one single processor individually. Interest in image mosaicking has been spurred by a wide variety of research and management needs. Parallel processing in remote sencing data processing free download as powerpoint presentation. A distributed parallel algorithm based on lowrank and. This parallel algorithm requires extra communication time, where the inef. An efficient framework for remote sensing parallel processing. Remote sensing rs data processing is characterized by massive remote sensing images and increasing amount of algorithms of higher complexity. Covers remote sensing data integration across distributed data centers. Parallel processing architecture of remotely sensed. The main contribution of hpgfs is that it introduces application aware data layout policies together with rs data object based data parallel io interfaces to provide direct and efficient accessing of rs data for various remote sensing applications. With sars contribution to remote sensing now wellestablished, the processing throughout demand has steadily increased with each new mission. High performance computing in remote sensing introduces the most recent advances in the incorporation of the highperformance computing hpc paradigm in remote sensing missions. Research on a kind of web based distributed forest remote.
Eighteen wellselected and organized chapters cover the entire spectrum of current and future data processing techniques in remote sensing. The rapid processing of mass remote sensing data put challenges on computers processing capability. Parallel processing of massive remote sensing images in a gpu architecture 199 using gpu to accelerate processing for all kinds of remote sensing, image pro cessing algorithms have resulted in greater related research achievements. Remote sensing image processing is characterized with features of massive data processing, intensive computation, and complex processing algorithms.
Mar 10, 2015 remote sensing applications it is a software application that processes remote sensing data. Parallel programing templates for remote sensing image. Pdf remote sensing data parallel processing base on. Parallel programming for dataintensive applications like massive remote sensing image processing on parallel systems is bound to be especially trivial and challenging.
Mapreduce has been widely used in hadoop for parallel processing largerscale data for the last decade. Pansharpening algorithms are dataand computationintensive, and the processing performance can be poor if common serial processing techniques are. Big data and high performance computing in earth system models esms are receiving increased attention in earth science research. The cloud has emerged as a promising new approach for adhoc parallel processing in the big data era because it is capable of accommodating various largescale platforms for different research areas with elastic system scaling. Remote sensing applications which may be addressed by neural. As an alternative to using hpc computing frameworks, other researchers have developed their own bespoke parallel largescale remote sensing data processing platforms. Distributed computing approach for remote sensing data.
A parallel processing algorithm for remote sensing. Parallel kmeans clustering of remote sensing images based. After a given data block has been processed, the results are merged on the master node. Grid computing, parallel computing, remote sensing, image processing 1.
Mining hidden information from rsbd for different applications imposes significant computational challenges. The autonomic model in remote sensing data processing system. The traditional centralized single mode becomes a bottleneck of remote sensing image processing which cannot meet the needs of future remote sensing image processing development. Currently, only a few parallel processing strategies are available in this research area, and most of them assume homogeneity in the underlying computing platform. Request pdf remote sensing data fusion algorithms with parallel computing the advances in satellite technologies, image analysis techniques and computational power make possible processing. The parallel and cloud computing platforms are considered a better solution for big data mining. Computer analysis of such remotely sensed earth resources data has many applications in agriculture,forestry etc. Grid computing, parallel computing, remote sensing, image processing. A free powerpoint ppt presentation displayed as a flash slide show on id.
Clustering is an important data mining technique widely used in processing and analyzing remote sensing imagery. High performance computing, cloud computing, and remote. Remote sensing big data management in a parallel file system. A parallel computing system of remote sensing images based on. Cloud computing has been identified as a potential solution to address some of the big data challenges in remote sensing,, and big data computing,, such as allowing massive remote sensing data storage and complex data processing, providing ondemand services,, and improving the timeliness of remote sensing information service. It focuses on the computational complexity of algorithms that are designed for parallel computing and processing. Therefore, high performance computing hpc methods have been presented to accelerate sar imaging, especially the gpu based methods. Pdf epub neurocomputation in remote sensing data analysis pp 262279 cite as. Parallel computing implementation of sar processing algorithms is therefore an important means of attaining high sar data processing throughput to keep up with the ever increasing science demand. Several studies have attempted to apply parallel computing to improve image mosaicking algorithms and to speed up calculation. Accelerating spaceborne sar imaging using multiple cpugpu. High performance computing, cloud computing, and remote sensing big data lizhe wang institute of remote sensing and digital earth radi chinese academy of sciences cas email.
Solutions for timecritical remote sensing applications the recent use of latestgeneration sensors in airborne and satellite platforms is producing a nearly continual stream of highdimensional data, which, in turn, is creating new processing challenges. The analyzed result shows that parallel computing cluster based on mpich can efficiently improve the speed of remote sensing image processing in the case of more complex algorithms. Frequently, processing for this type of images has faced to some difficult issues, such as calculating slowly or consuming huge in resources. We presented a clusterbased parallel processing system for hj1 satellites data, named cluster pro. In order to obtain the ground information more accurately from the images, the remote sensing image processing would have several steps in the aim of better image restore and. The kmeans clustering is a basic method in analyzing rs remote sensing images, which generates a direct overview of objects. A parallel file system with applicationaware data layout.
In particular, i focus on a problem that can be decomposed into a set of independent tasks, which on a serial computer would be performed sequentially, but with a cluster computer can be performed in parallel, giving an obvious speedup. Remote sensing data fusion algorithms with parallel computing. Geospatial data are one of the most significant types of big data, and the rapid growth of such data has imposed enormous challenges to current methodologies, applications, and infrastructures 1,2. Parallel kmeans clustering of remote sensing images based on mapreduce 163. Aim at the disadvantages of super computer and cluster parallel image processing system.
Through parallel programming environment based on message passing, parallel kmeans unsupervised classification of remote sensing image with different sizes we performed in parallel environment with different computers number. However, conventional clustering algorithms are designed for. It is useful in the field of remote sensing analysis as well, where objects with similar spectrum values. It describes the entire life cycle of remote sensing data and builds an entire high performance remote sensing data processing system framework.
To improve the efficiency of this algorithm, many variants have been developed. These characteristics make the rapid processing of remote sensing images very difficult and inefficient. Generic parallel programming for massive remote sensing. The goal of this research is to put up web based distributed remote sensing parallel processing services for forest researchers and educators.
Remote sensing free fulltext a parallel computing paradigm. A distributed parallel algorithm based on lowrank and sparse representation for anomaly detection in. Remote sensing big data rsbd is generally characterized by huge volumes, diversity, and high dimensionality. The control data cor poration flexible processor system is a commercially available multiprocessor sys tem which has been recommended for use in remote sensing 4,5. Covers remote sensing cloud computing covers remote sensing. Parallel computing rendering in specific remote sensing image. Through the analysis of the cloud storage, cloud computing technology and characteristics of the remote sensing data, a method of mass remote sensing data stored in the cloud is put forward in this paper. Parallel kmeans clustering of remote sensing images based on. A parallel processing algorithm for remote sensing classification j. The method grounds on two principles, parallel computing and digital photogrammetry. Cloud computing offers the potential to tackle massive data processing workloads by means of its.
Experiments indicated that gspcs has the capability of processing remote sensing images effectively in parallel. Soft computing for environmental applications and remote. Pdf inmemory parallel processing of massive remotely. However, for largescale applications, remote sensing image mosaicking usually requires significant computational capabilities. Nov 03, 2010 parallel computing rendering in specific remote sensing image processing therefore, remote sensing images always have features with multidimensions details and huge size. In remote sensing applications, there is an increasing need to analyse remote sensing data over large areas, which involves mosaicking massive remote sensing images. Introduction recently remote sensing systems have improved in several ways that require new computational power to fully exploit them. Parallel computing execution of several activities at the same time. Traditionally, a cpubased cluster or cluster for short computing system with distributed memory is the prevailing architecture of the modern hpc technique, and has been widely used for parallel. Geoanalysis with remote sensing images is often conducted under the condition that the images are radiometrically consistent. The computing nodes could run parallel algorithms separately to achieve more finegrain parallelism. Huang has been the lead chair for the spie conference on satellite data compression, communications, and processing since 2005, for the spie europe conference on high performance computing in remote sensing since 2011, and for the ieee international workshop on parallel and distributed computing in remote sensing, in conjunction with the. The proliferation of rs data is revolutionizing the way in which rs data are processed and understood.
1472 1335 370 1485 1261 940 699 1253 72 110 206 709 23 1457 1027 1251 1191 1660 339 1377 1611 346 1007 383 924 1155 307 760 949 1671 841 1656 46 47 1229 1314 1183 910 748 979 410 286 726