The main task of image fusion is to extract the advantage information from the source images. Multiscale transform (MST) is commonly used in image fusion, but it is difficult to integrate detail information and structural information from source images. Therefore, it is important to integrate the information in each image patch group. Different from MST, edge preserving technology obtains more edge information than MST by extracting spatial information. Based on this superior feature, we propose an infrared and visible image fusion method via rolling guidance filter (RGF) and weight map. First, source images are decomposed into multiscale detail layers and structural base layers by RGF; second, to avoid the loss of local information, the modified Laplacian energy weight matrix is used as detail weight maps to merging the detail layers; then to reduce the loss of contours of edge objects, a structural weight maps are used to merging the base layers; finally, these two different weight maps are applied to reconstruct the fused image. Simulation results indicate that our fusion approach performs better in merging detail information and structural information than some classic MST fusion algorithms and the latest fusion methods.
Generally, 2-D spatial data are divided as a series of tiles according to the plane grid. To satisfy the effect of vision, the
tiles in the query window including the view point would be displayed quickly at the screen. Aiming at the performance
difference of real storage devices, we propose a 2-D tiles declustering method based on virtual device. Firstly, we
construct a group of virtual devices which have same storage performance and non-limited capacity, then distribute the
tiles into M virtual devices according to the query window of 2-D tiles. Secondly, we equably map the tiles in M virtual
devices into M equidistant intervals in [0, 1) using pseudo-random number generator. Finally, we devide [0, 1) into M
intervals according to the tiles distribution percentage of every real storage device, and distribute the tiles in each interval
in the corresponding real storage device. We have designed and realized a prototype GlobeSIGht, and give some related
test results. The results show that the average response time of each tile in the query window including the view point
using 2-D tiles declustering method based on virtual device is more efficient than using other methods.
It is the tendency for the development of massive spatial data network service to use cluster to enlarge load capacity of
spatial data server. In this paper, we use the OSD (Object-based Storage Device) storage cluster as the shared storage of
LVS (Linux Virtual Server) server cluster, and use the servers in the server pool of the LVS server cluster as the storage
client of the OSD storage cluster, to build a scalable massive spatial data network service architecture, which uses the
high scalability of the LVS server cluster and the OSD storage cluster to avoid the bottlenecks of massive spatial data
network service bandwidth and storage I/O throughput.
Several load balance scheduling algorithms embedded in the LVS server cluster can satisfy the demand of load balance
in many applications. But those algorithms can't optimize load balance of spatial data servers, regardless of the features
of spatial data. Spatial data in large scale network service application is generally organized according to the global
longitude and latitude, and managed according to the principle "vertical hierarchies and horizontal dividing". According
to the features of spatial data, we optimize the scheduling algorithm to enhance the Cache utilization efficiency for single
spatial data server.
With the rapid increase in the amount of spatial data and the number of geospace information
system users, current network architecture of geospace information system mainly based on a single
server has two primary data transport bottlenecks: the bottleneck of network services throughout
provided by the server and the bottleneck of I/O throughout of the storage system for distributed spatial
data. In order to avoid both bottlenecks, we present a C/S mode geospace information system based on
double-cluster, namely server cluster and storage cluster, and it is called GlobeSIGht.
In GlobeSIGht, we use a Linux Virtual Server (LVS) cluster to avoid the bottleneck of network
services throughout provided by the server, and use an Object-Based Storage (OBS) cluster to avoid
the bottleneck of I/O throughout of the storage system for distributed spatial data. Spatial data is
organized as spatial storage object stored in the object-based storage device, and the metadata server
manages all the object-based storage devices and spatial storage objects to provide a spatial data shared
storage pool as the backend storage of the LVS cluster. Furthermore, by using plug-in technology,
GlobeSIGht has integrated many relevant application systems such as 2D geographic information
system, 3D geographic information system, multimedia application subsystem, massive MODIS
remote sensing images management and dispensing online system.
It is very difficult to design an integrated storage solution for distributed remote sensing images to offer high performance network storage services and secure data sharing across platforms using current network storage models such as direct attached storage, network attached storage and storage area network. Object-based storage, as new generation network storage technology emerged recently, separates the data path, the control path and the management path, which solves the bottleneck problem of metadata existed in traditional storage models, and has the characteristics of parallel data access, data sharing across platforms, intelligence of storage devices and security of data access. We use the object-based storage in the storage management of remote sensing images to construct an object-based storage model for distributed remote sensing images. In the storage model, remote sensing images are organized as remote sensing objects stored in the object-based storage devices. According to the storage model, we present the architecture of a distributed remote sensing images application system based on object-based storage, and give some test results about the write performance comparison of traditional network storage model and object-based storage model.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.