Contrast Stretching

来源: https://vis-webcontent.s3.amazonaws.com/quickguides/Getting+Started/Contrast+Stretching/index.html

Lesson content

Lesson 1 of 3

About Contrast Stretching

Image pixels can have any value: negative, positive, integer, or floating-point. When a single band of image data is projected on a screen, pixels are displayed as brightness values, or digital numbers (DNs). A data pixel with a larger value is brighterthan one with a smallervalue. However, unlike the image data, typical monitor pixels can only have 256 unique brightness values. This limitation prevents most of the data from being displayed with a brightness exactly equal to their real value.

As its name implies, contrast stretching is a display technique used to improve the contrast of an image. It can help to emphasize and delineate features in an image.

Contrast stretching utilizes the full dynamic range of the raster by stretching the DNs to a range that can be displayed on a digital screen, namely from 0 to 255. For example, if DNs range from -1.0 to 1.0 for a spectral index image, a linear 2% stretch attempts to visually remove very dark or very bright pixels by trimming the lower 2% and upper 2% of the histogram. All intermediate values are assigned brightness values linearly.

Features of interest may only occupy a small percentage of the full dynamic range. Contrast stretching allows you to focus on specific parts of the image histogram to better interpret certain features. Different techniques are used for contrast stretching such as linear, linear percentage, equalization, and Gaussian. You will learn more about these in the next hands-on exercise.

CONTINUE