IEEE Geoscience and Remote Sensing Magazine - March 2013 - 48

Image and
Information Files
Research Topics
Image
Processing
Research &
Teaching

Search &
Requests

New Algorithms
& Techniques

NASA Scientist
* Domain
- Earth Science
- Planetary Science
- Exploration
* Image Data
- Location
- Format
- Size
* Application
- Processing Type
- Desired Result/
Truth Data

NASA
Needs &
Challenges

Image Data, Information,
& Validation Data

Data Pre-Processing and
Database Populating

Client Browser

Request Data

Requests
Web and
Application
Server
Replies

Database
Server
Send Data

FIGURE 2. Overall organization of the IMAGESEER database.

C. IMAGESEER Content
As stated previously, the current
The first IMAGESEER
version of IMAGESEER focuses
prototype includes
on Earth Science, although the
a representative
database and the website are
built to accommodate other
sampling of NASA
domains such as Planetary Scimultispectral and
ence and Exploration. Workhyperspectral images
ing with project scientists and
from several Earth
engineers, IMAGESEER identiScience instruments.
fied Earth Science applications
of interest and their corresponding image processing
techniques with accompanying data, data types and database requirements.
More specifically, we performed a preliminary identification of:
◗◗ The types of image processing problems to use for training/
teaching, namely, we identified 4 techniques: 1) Map
Cover/Classification, 2) Cloud Detection, 3) Image Registration, and 4) Gap Filling/In-painting
◗◗ The regions of interest. We identified 4 regions of interest corresponding to different geographic features,
Mountainous area/Colorado, Urban area/Los Angeles,
48

Water-Coastal area/Chesapeake Bay, and Agriculture
area/Quincy (Illinois)
◗◗ The instruments of interest: Landsat-4&-5/Thematic
Mapper (TM), Landsat-7/Enhanced Thematic Mapper
(ETM+), Earth Observing-1/Advanced Land Imager
(ALI) and Hyperion, MODIS and SeaWIFS
◗◗ The datasets sizes: preview (thumbnail) images, 1K#1K
and almost full scenes
◗◗ The datasets formats: TIFF and Raw data
◗◗ The security and privilege related issues (e.g., related to
courses or exercises proposed by the teachers vs. general
access by professors, students and researchers).
The database was designed and implemented following the above requirements. Data was gathered and
pre-processed for four of the instruments of interest;
Landsat-Thematic Mapper (TM) and Enhanced Thematic
Mapper (ETM+), EO-1/Advanced Land Imager (ALI) and
EO1/Hyperion. For all datasets and all image processing
techniques, truth data was identified and all data was
ingested into the database; classification ground truth was
extracted from two National Land Cover Data (NLCD)
maps gathered in 1992 and 2001 by the Multi-Resolution
Land Characteristics Consortium (MRLC) [13], cloud
cover maps are computed using the software available
ieee Geoscience and remote sensing magazine

march 2013
Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2013
