Image processing is a group of methods, all of which change the way an image looks. These techniques are useful in many ways. You can correct images that are too dark or too light, that lack contrast, or are washed out. You can also create special effects with image enhancement.
GIFConverter supports some image enhancements that fall into the category of point processes. These are effects that take place individually for each pixel (point) in the image. Changing the brightness is an example of a point process.
Stretching and rotation, mentioned earlier in this manual, are examples of a geometric process. A geometric process changes the position or arrangement of pixels in the image.
There are two kinds of processing that GIFConverter does not do: area and frame processing. An area process changes a pixel’s value based on its own value and the value of neighboring pixels. One example of an area process is sharpening, which enhances edges. A frame process calculates a new image that is a combination of two previous images.
Currently, GIFConverter only allows image processing on images with 256 colors or less. To do image processing on full color images, see “ Reducing the Number of Colors ” for more information
A gray map is a function that assigns a new intensity to a pixel based on that pixel’s current intensity. A pixel of the lowest intensity is black, and a pixel of the highest intensity is white. A pixel of intermediate intensity may be gray or have a color. As a bright color is moved toward white or black, it loses saturation, or colorfulness.
An example of a gray map.
The example above is a gray map that makes no changes to the image. For each intensity level on the input, the curve defines an output intensity level (dashed line).
A histogram is a graph that shows how much of the image has a certain intensity. Here is an example of a histogram:
An Example Histogram.
The height of the histogram at any particular point shows the number of pixels or area of the image that has that intensity. The output side has no units. This is because the relative values are more important than any absolute value. In this case, the image is made up of mostly dark values.
Histogram equalization is a technique that generates a gray map which changes the histogram of an image to be as close as possible to a user-specified desired histogram. This technique is useful for processing images that have little contrast. Typically, the desired histogram is a level one, meaning that all intensities are equally represented in the image. The net effect is to turn an image with details imperceptible to the eye into an image that is easily seen.
Gamma is a measure of the nonlinearity of a monitor. This means that the RGB values we send to the monitor do not correspond to the intensity we see in a linear fashion. The intensity seen on a monitor can be specified as:
I=Vγ
I is the intensity seen on the screen, V is the value from the image, and γ is the gamma.
The Macintosh is set up so that the gamma is 1.4.
Here are gray maps representing a gamma less than one, and a gamma greater than one:
 |
 |
|
Gamma
< 1
|
Gamma
> 1
|
Some images may appear washed out. This is because the values in the image have been corrected for systems that have a gamma greater than 1.4. Most PCs used with common monitors have a system gamma of 2.2.
GIFConverter’s internal image processing apparatus looks like this:
GIFConverter takes the following steps in applying image enhancement:
You are able to invoke all the image enhancement features through the Enhance... command on the Image menu, which produces the following dialog box.
To enhance images, first select the images with the pointer tool. Image enhancement does not work on portions of images. You can only enhance entire images.
The Enhance dialog box consists of two main areas:
The Display shows data at a specific point in the image enhancement process described in the previous section. When viewing the Desired Histogram or User-Specified Gray Map modes, clicking and dragging in the display changes the histogram or gray map.
To the right of the display are several controls that change the way GIFConverter displays an image.
When you click on the OK button in the Enhance dialog box, GIFConverter applies the settings to all the images that you selected.
The large area at the left of the dialog box is the display area. Above the display is the Display menu. This menu selects the data that shows in the display area.
Click on the display menu to choose among the options:
The Gamma slider allows you to change the input characteristics of GIFConverter’s image display system. When viewing a washed out image, adjust the gamma to a value in the range of 1.5 to 2.2.
Sometimes, when GIFConverter reads an image, it comes out negative. Some TIFF files do not contain an explicit specification of how to represent black and white values. GIFConverter makes an assumption about this representation according to the TIFF specifications, but images from some programs still appear inverted.
Check the Invert Image button in the dialog box to cause GIFConverter to reverse the image values. This will create a negative of the image, which will correct for images that start as negatives.
Histogram Equalization is a technique for increasing the detail of an image that is lacking in contrast. This technique changes the intensity levels in the image to cause the image to conform to some desired histogram. Histogram Equalization helps the quality of dithered images in the One Bit mode. Sometimes, this technique increases the contrast too much.
To turn this feature on, check the Histogram Equalization check box.
The default histogram contains equal amounts at all intensities. To change the histogram, select Desired Histogram from the Display menu in the dialog box. Then draw the desired histogram into the display box with the mouse. Press the Reset button to return to the default histogram.
The Contrast and Brightness controls operate similarly to the controls on a TV set.
The Contrast control is measured as a percentage. 100% corresponds to normal contrast; lower values indicate less contrast and higher values indicate more contrast. The contrast can be adjusted from 25% to 400%.
The Brightness setting is an intensity value that is added to all values in the image.
The user-specified gray map offers the opportunity to apply unusual changes to the image.
To alter this gray map, select User-Specified Gray Map from the Display menu in the dialog box. Draw the desired gray map into the display area. Click on the Reset button to restore the default gray map.
When you turn off Automatic Dithering , you will not see your changes immediately. After performing an image enhancement, choose the Recalculate command from the Image menu to update the screen.
Copyright ©1998 by Kevin A. Mitchell. All rights reserved.