|
Page 3 of 3 What is a Digital Image?
A digital image is:
- a rectangular array of measurements (usually of light, but they may be measurements of things such as temperature, x-rays, the elevation of points on the surface of a planet, etc.) sampled at regular intervals and rounded off to the nearest whole number; and
- displayed according to a color Lookup Table (LUT), which translates each measurement into a pixel value that is then assigned a specific color or shade of gray.
 In other words, a digital image is a long string of numbers representing different levels of brightness for grayscale images, and brightness and color for color images (such as RGB images). When the numbers are placed in the correct order (rows and columns), and a shade of gray or a color is assigned to each number by an LUT, a recognizable picture is formed. Look at the series of magnifications of the Antarctic ozone hole on the right to see how a digital image is really an array of numbers mapped to colors.
Pixels also correspond to distances and areas. For example, a length of twenty pixels in an image may represent one micron in an electron micrograph or one hundred kilometers in a satellite image. The more pixels there are in an image, the greater its resolution; an image with a high number of pixels and a high bit depth therefore more closely represents the actual thing that was measured. Bit depth tells you how much data is being used to define each pixel in an image. A bit is the smallest unit of data, and comes in two possible states: either 1 or 0 (or on/off, or black/white). A black-and-white image is a 1-bit image, since only one bit is needed to define each pixel; 8-bit and higher images give you a lot more information. Unlike graphics, where meaning comes only from the patterns formed by many pixels, each pixel in a digital image may contain meaningful information.
What is Image Processing and Analysis?
Image processing is the manipulation of numeric data contained in a digital image for the purpose of improving or altering its visual appearance, either for aesthetic reasons or to make image features easier to see. This may be done by using image processing functions such as brightness and contrast enhancement or application of filters that reduce noise (irrelevant data) or sharpen edges. LUTs can be used to apply color to a grayscale image or to redefine the values of pixels in a color image so that certain types of image information become more obvious. For example, the pixels representing a particular type of ground cover or weather phenomenon shown in a satellite photo can be displayed as one bright color.
A major purpose of image processing is to alter the appearance of features in an image so that they may be better measured. Image analysis involves collecting data from digital images in the form of measurements (e.g., measuring the size of a tumor in an x-ray of a brain, or determining the amount of sulfur dioxide gas and ash in the plume of a volcano shown in a satellite photo). The accuracy and precision of measurements using digital image analysis is far greater than is possible with rulers and calipers.
Travel Within Data: The Power of Digital Images
Digital images are powerful because they are composed of thousands of pixels representing data that can be displayed in virtually any color or shade using computer software such as NASA Image2009. An image can be manipulated to bring out features not apparent at first glance, complex structures can be accurately selected and measured in many different ways, and time-lapse sequences can be animated.
Image processing and analysis allows learners to examine scientific information in ways that are not possible with textbooks, videos, interactive multimedia, and Internet Web pages. While other tools can show static images of structures, play animations, or display three-dimensional representations of objects, only image processing and analysis permits learners to travel within data and conduct scientific investigations on what they are visualizing.
|
