Ward, Cal and Matt Betts

There is no such thing as a perfect picture(image) within image enhancement. All images are subject to individual wants , needs. desires and interpretation of each individual image.

Images are computer representations of pictures displayed on a computer monitor or screen. Image enhancement is a process of remote sensing, used to improve visual analysis and make certain desired aspects more interpretable to the human eye. Because the computer recognizes the image only as digital representations of actual landforms and features, that representation can be altered, or enhanced, in a variety of ways to bring out designated areas or certain aspects of an image.

For our project for Remote Sensing II (Geog 576), we decided to write a remote sensing tutorial for Erdas Imagine. Our purpose is to write a tutorial that will benefit people who have no prior experience with Erdas Imagine. We focussed on explaining the various image enhancement routines. Image enhancement routines are mathematical algorithms used to improve visibility in remotely sensed images and make them more interpretable to the human eye. Our intention is to provide easily understood information and examples that will help the beginner to understand the benefits and processes used by image enhancement algorithms.

We used three steps to approach this goal:

  • Explain the idea behind each image enhancement routine. In this step we tried to define what the purpose and goal of the specific image enhancement routine.

  • Next we defined the mathematical processes behind each enhancement routine.

  • And third we demonstrated the usefulness of each algorithm with examples from the San Rafael Swell area in Emery County, Utah.

    For our project, we used the Erdas Imagine Field Guide (3rd edition, 1995), and Introductory Digital Image Processing (John R. Jensen, 2nd edition, 1996) to help us understand the processes behind the routines. We also used the Help and View functions for a lot of the .gifs and for help understanding the underlying processes.

    This project we intended to explain the image enhancement routines for spatial enhancement, radiometric enhancement and spectral enhancement in Erdas Imagine. But due to time constraints and difficulty finding a weekend when the server wasn't down, we cut the scope of our project.

    We will be using an area of Utah known as the San Rafael Swell located in Emery County. This is the dark, inverted horseshoe shaped area below the notch cut out of right corner of the state.

    This is our original TM image from the San Rafael Swell.

  • Spectral Enhancement

    After Imagine is open you can get to the spectral enhancement routines by selecting the interpreter icon at the top of the Imagine command window. When you select this icon you will get a Image Interpreter window.

    Spectral Enhancement is the third choice on this list. After selecting this button it will bring up the following window.

    Spectral Enhancement is the changing of the values of each pixels in the original image by transforming the values of each pixel along a multiband basis. Spectral Enhancement allows different features that have specific reflective characteristics in different bands of the electromagnetic spectrum to be compressed if data is similar . It also allows modifying of the pixels of an image independent of the values of surrounding pixels.

    Spectral enhancement creates new bands of data that are more interpretable to the eye.

    Spectral Enhancement which has principal component as the first choice. Selecting this button will bring up the Princilpe Component as the first choice. Selecting this button will bring up the Principle Component selection box.

    In the input file you must select an image which has been previously saved, by selecting the icon which looks like a file folder to the right of the input box. Then you must type in the new name in the output box.

  • Principle Component This Enhancement routine looks for redundant data recorded by the different bands and compresses it grouping that data together. The reduction in the number of bands often allows tthe user easier visual interpretation.

  • Inverse Principle Component We did not complete this Enhancement routine.

  • Decorrelation Stretch This Enhancement routine applies a linear contrast stretch to the Principal Components of an image.

  • Tasseled Cap

  • RGB to IHS This Enhancement routine transforms the image from the standard red, green and blue values to an alternative color space with values corresponding to intensity, hue and saturation.

  • Indices

    The Indices are mathematical equations that are used to compress the multiple bands of data down to just a very few band or even just one band that contains the information desired. These mathematical equations are based on surface reflectance and absorption properties which are in turn based on the chemical composition of the surface. A wide variety of on-ground characteristics can be evaluated by using these equations. Things such as vegetation health, biomass, mineral types in the soil, vegetation types, leaf area index (LAI), amount of radiation in the photosynthetic bands that are being consumed, and percent ground cover vegetation.

    These preprogrammed indices are usually very simple and have a great deal of useful information to convey.


    ERDAS, 1995, EDRDAS Field Guide, 3rd Atlanta: ERDAS, Inc.

    Jensen, J. R., 1996, Introductory Digital Image Processing, 2nd ed., NJ: Simon & Schuster.