Document Actions
Scientific Analysis
MDA scientists develop the appropriate algorithms and ensure that they work accurately. MDA software engineers build the algorithms into a reusable application that runs quickly and is user-friendly.
Remote sensing technology today has advanced to the point that it is able to quantitatively answer demanding scientific challenges. The Image Processing Division approaches the development of solutions in two steps. First, a scientist well-versed in the problem at hand develops the appropriate algorithms and ensures that they work accurately. Then, software engineers build the algorithms into a package that runs quickly and is user-friendly so that it may be used in the future for other applications. The Image Processing Division is staffed by a diverse group of scientists ranging from computer scientists to geologists to engineers to geographers.
Proprietary Software Development
MDA has been developing its own powerful Image Processing System (IPS) since its inception in 1969. Written in C and C++, IPS performs many tasks faster and better than COTS packages, and has many unique features. Other software development includes a graphical user interface (GUI) and production pipeline for Landsat 7 ETM+ images using the LPGS code distributed by USGS. Many algorithms are made specifically for scientific applications desired by the client. Most of the hyperspectral image processing algorithms developed at MDA are written in IDL and are easily integrated within the ENVI.
Application Development
Many of the algorithms we develop are in direct response to our clients needs, while others are for general use. Below are several broad examples of applications we have developed algorithms for. Contact us to discuss how we can develop solutions to your unique challenges.
Lithologic Mapping
Passive remote sensing methods only see the top few microns of any land surface, while active methods can only penetrate a few meters at most. But by mapping the lithology at the surface, geologists can interpolate the subsurface structures. MDA has developed multiple tools for use with multi and hyperspectral data that can map down to a subfacies level.
Hydrocarbon Seep Mapping
Natural subsurface hydrocarbon traps nearly always penetrated in some manner and release liquids and gases that migrate to the surface. Most of the time it is in such a low volume that no effect can be visibly seen. However, the geochemical changes these hydrocarbons cause produces a characteristic set of minerals and vegetation response that can be mapped with remotely sensed imagery. We have developed algorithms to detect these subtle spectral changes in environments ranging from arid to tropical.
Vegetation Stress Mapping
Hyperspectral data allows the unprecedented ability to accurately measure biophysical parameters and to use these to map vegetation stress. Development is underway in bioengineering and image processing to use spectral information to track down the cause of the vegetation stress. This application has a great number of uses, including agricultural monitoring, environmental monitoring, and natural hydrocarbon seep detection.
Ground Truthing
Ground truthing is an often overlooked aspect of remote sensing. After all, isn't remote sensing intended to replace the need to go visit the area of interest in person? The answer for the most part is yes, but only once it is a certainty that the information gleaned from the imagery is accurately and truthfully representing what is happening on the ground. MDA scientists have traveled all over the globe to ascertain that what we derive from the imagery is accurate. The use of field portable spectrometers is important to characterize materials fully and to ensure that atmospheric corrections are done properly.
