2.1 Creating a histogram of the whole image |
||||
Selecting the whole image Creating a histogram The histogram window |
||||
Images: atsr199701.gif atsr199801.gif Description Download images (195 K)
Useful information: |
Histograms contains information about the distribution of pixel values within an image area. Creating histograms in Bilko involves two steps:
Note: Histograms are also known as bar charts because they consist of a series of bars where each bar represents a range (or class)of data values (in 8-bit data each integer value has its own bar). The height of each bar is proportional to the frequency of the class it represents (i.e. proportional to the number of times a value in that range occurs within the image area selected). Selecting the whole imageFirst activate the image you want to study, in this case atsr199701.gif. (If you have closed this, please open it again, as described in section 1.1). The easiest way to select the whole image is to use Select All. As with most actions in Bilko, there are a number of ways to do this:-
From the menu bar:
Note: You could click the box selector on the tool bar, point your mouse at the top left corner, and drag it to include the whole image, but it's easy to miss rows or columns along the edges, so the methods above are the ones we recommend. Creating a histogramHaving made your selection, you can now open a new histogram doucment corresponding to this area.
From the menu bar:
This opens te New dialog with HISTOGRAM documents highlighted. ( figure (5K)). Make sure the Apply Stretches box is unchecked before pressing OK to accept. Note: If histogram is not one of the options listed in the New dialog, this is because no area was selected. Activate the image and try again, using [CTRL+A] to select the whole image. The histogram window
A histogram window will appear. This shows the number of pixels in the selected image having a given pixel value (y-axis)
versus the set of values (DN) that pixels can take (x-axis). You will probably notice that in this particular histogram
the scaling of the y-axis scaling is not ideal (see
figure
(5K)),
but makes it hard to see the frequency distribution of pixels with values of less than 255.
|
|||
Answers: |
Question 1. In the next section you will learn how to make sure the histogram gives a valid representation of the true data values in the selected image. |
|||
|