Email: proof@proofofconcepts.com

The histogram of an image is a statistical graph showing the number of shades of gray (DN/pixel value) and the frequency at which they occur in the image. Open the Florida image (image2.tif.sit) in the NIH Image program. If the program gives you two images one titled Red and the other Indexed Color. Close the one titled Red.

Go to the menu bar, click on Analyze and then click on Show Histogram. You will open an image which looks like this - this is a histogram. Notice that it shows a group of lines on the left side of the histogram window. Histograms are displayed as a bar graph, with each shade of gray represented by one of those vertical lines. The height of the bar displays the number of pixels with that value. Scroll the cursor across those lines and look at the changing numbers in the info box.

The word Level (pixel value) tells you the gray scale number (DN) value and the Count tells you the quantity of pixels in that image which have that value. Therefore if your mouse hair is at Level 24 you should see a pixel count of 7743. Knowing this is important because knowing the brightness value of an image and the number of pixels at that level can help you in determining many physical characteristics of the features in the image.

Determine the number of pixels for 5 different (DN) values in this image - your choice.

This image is fairly light and shows little detail. In fact the shades of gray range from a DN of 0 to a DN of 32.Considering that the darkest gray scale DN is 255, there are no pixel values darker than 32.

Applying the contrast and/or brightness controls to this image, changes the brightness values of the pixels but does not change the pixel values themselves. This activity will show you how to "stretch" the pixel values over a selected range of pixel values. Stretching these values will help to bring out features not readily apparent in the original scene.

In order to use the full range of pixel values, the maximum pixel value in this scene must be shifted to 255. This becomes a mathematical operation. What number would you have to multiply the maximum pixel value (32) by to get 255 (round to the nearest whole number)? Do this and apply the operation to the image by clicking on the Process/Arithmetic/Multiply menu. Insert the number you determined into the dialogue box and apply. What does the image and histogram look like now? Save the image with a title such as histoa.tif. Why do you suppose there are "blanks" between the vertical lines in this new histogram? What is the new pixel value for the 7743 pixels that were originally assigned a DN of 24? Close this image and re-open it again for the next activity.

Try stretching the pixels in the image to a maximum of 64. Save the image and e-mail it with the name histob.tif. Once again close and re-open the image. Try stretching the pixel values to a maximum of 50. Close and re-open once again. Use the Map controls to change the contrast and brightness of the image. Notice how the Map graph, image, and LUT change as the controls are changed. Look at the histogram. Have the pixel values changed?

By changing the contrast/brightness of the image or mathematically stretching the range of pixels, features could be seen which were previously not apparent. In the image you have been working, water areas such as rivers, streams, and lakes are enhanced and are more easily viewable. City areas and agricultural lands also stand out , along with different weather features.

Hurricane Opal Project Activity