318
To establish a image Pyramid has two main ways :the first one,
the multi-resolution data source automatically build the
pyramids; the other one , the bottom of the pyramid data is the
raw data, other layers of image data are collected from the
bottom to build the pyramids.
For the second method, it can be divided into two methods:
Using the ratio to build the pyramids, forming more resolution
levels. The other is based on a scale of the requirement to
conduct a scale of the corresponding image data.
3.3Visualization and query of image data
1) Visualization of image data
Image data’s display are the process of mapping a rectangular
region images on the ground to a screen window scope . Image
data in the database are geocoded, so it need establish
corresponding relationships between geographical coordinates
and screen coordinates during the visualization process. After
coordinate conversion, we can visualize the images. The
specific process is seen from figure 4.
2) Inquiry of image data
Inquiry is the basic functions of a database system , a good
query mechanism will provide users with convenient
information services ,it can enable users to extract information
that they are concerned quickly from the large number of data.
Inquiries of image data have spatial query, metadata inquiries,
map index inquiries, inquiries View directly, content-based
inquiries etc.
3.4 Update the image data
Because of detecting automatically updating technology
information-based image texture and structure information the
changes are still at the research stage. So currently adopt
artificial manner to update the data. Because raster data
modification and editing are very complex, it is commonly
used the data in the data storage modules to replace whole to
implement to update the data or database(used new data to
replace the old data). Generally we do not add, delete, modify
some local data in a data unit. All data which are used to be
updated are already finished beforehand ,update the data or
database is only a data replacement process.
Currently production modes of spatial data are maps, and the
situation is difficult to change. The USGS produce digital maps
and orthograph with maps as units. Therefore, we update the
data with maps as units. The new data are submitted as
maps to a database once again. Using this method to make the
data update has the virtue of low maintenance costs and high
security performance.
Fig4. Image data visualization algorithm flowchart
4. DESIGN AND BUILD THE IMAGE DATABASE
SYSTEM
This paper establishment a remote sensing image database
system according to a project request in a city. Currently, this
city’s image data is Quick Bird Panchromatic Remote sensing
image data. Quick Bird satellite remote sensing was
successfully launched on October 18, 2001 by the United
States of Digital Earth (Digital Globe) company ,which is the
current satellite remote sensing data acquisition highest level,
its panchromatic image’s spatial resolution radio has reached
0.61 meters. It is the highest-resolution radio commercial
remote sensing images in the world currently. Quick Bird
shoots more extensive area, compares with aviation
Photogrammetry. Each image is nearly 300 square kilometers,
cost is low, full color image data required only 20 dollar per
square kilometer, mapping cycle is even shorter. In this paper,
we use Panchromatic Remote sensing image data of a city
Quick Bird which shot in 2006, adopt Visual C + + 6.0-based
Oracle 9i city high-resolution remote sensing image database
system. Introduce original image data and extract
corresponding scale image data to establish the three-tier
pyramid structure with using total and child database as the
logical structure of image data.
4.1 The basic function of image database
A image database system on a city was built in this paper. Its
main functions are the following :
(1) Create image database. Create image project total and
child database.
(2) Input image data to the database. Single color images and
a number of color images store function automatically.