HOW TO

Understand general information about images

Last Published: April 25, 2020

Procedure

Summary

This document provides general information on images. It includes a description of the type of information held in image files, the difference between single band and multiple band images, and the difference between pseudocolor and true color images.

Discussion

An image is a binary file of unsigned integers. It does not contain floating point or negative values. An image has header information showing the number of rows and columns. The header may also give georeferencing and/or statistical information, but this is not the case for all image types. Header information may be stored within the image file or in a separate file.

An image may have one or many bands. Each band holds a different type of information. A basic example is an image that holds several different spectral ranges of light. One band can hold the red component of light, another the green, and the last the blue. Thus, each pixel in the image will hold three pieces of data for the same location. Both single band and multiple band images are stored in one file (this refers to actual data). The header and colormap information may be stored in separate files.

A one band image may represent black and white, grayscale, or color data. Black and white images use 1 bit of information for each pixel. The value of the background is usually 0 and the foreground is usually 1. Grayscale images
use from 2 to 32 bits per pixel.

One band color images are called pseudocolor images. They have 4 or 8 bits of data per pixel, along with a colormap. The colormap contains an index number, and red, green, and blue values that make up the color for that index. They are called pseudocolor images because they are limited to a total of 256 (in the case of 8 bits) colors out of the total spectrum. The colormap may be contained within the image file, or it may be in a separate file.

Multiple band images may be displayed as grayscale or color. Each band of a multiple band image usually contains 8 bits of data per pixel. One band can be displayed as a grayscale image or a combination of any three bands as a color image. When three bands are displayed as color, it is called a 'true color' image. The three bands are used to give a value of red, green, and blue for each pixel. The color resulting from that combination is the color of the pixel. These are called true color images because they can show up to 2^24 or 16,777,216 different colors, which is more than the human eye can distinguish.

Note:
For more information, see ArcDoc > Image Integration > Introducing the Image Integrator.

Article ID:000001272

Software:
  • ArcMap 8 x
  • Legacy Products

Get help from ArcGIS experts

Contact technical support

Download the Esri Support App

Go to download options

Discover more on this topic