THE MAGAZINE FOR COMPUTER APPLICATIONS
Circuit Cellar Online offers articles illustrating creative solutions
and unique applications through complete projects, practical
tutorials, and useful design techniques.

IMAGE PROGRESSING FUNDAMENTALS

Part 1: Playing with the Pixels

by James Antonakos

Start ý Imaging 101 ý RLE Compressor Tool ý Adjusting the X-Y Resolution ý Adjusting the Z-Axis Resolution ý Simple Image Operations ý Run-Length Compression ý More to Come ý Sources and PDF

SIMPLE IMAGE OPERATIONS

Table 1 lists several basic image processing operations. Some operations alter the way pixels appear in the image, and others actually move the pixels around within the image.

Photos 4a and b show the histogram for the source image. There are two tall spikes, one for the numerous black pixels (the hair and background) in the image, and the other for the white pixels (the shirt). Photo 4b is the histogram for the same image after its z-axis has been adjusted to 3-bit pixel values (eight shades of intensity). Note that there are only eight spikes in the histogram, and they are all tall. This is because there are still over 65,000 pixels in the image, but with only eight different shades, each shade must account for a large number of pixels.

a)	b)
Photo 4aýThis histogram of the original image shows a lot of intensity variation. býThis is a histogram of the same image when 3-bit pixel values are used. One technique, called histogram equalization, is used to evenly distribute the shades of pixels in an image, improving the contrast.

The threshold operation is used to convert an image into a binary image. Each pixel in the image is compared against the threshold value and then replaced with one of two different shades, depending on whether the pixel value is larger or smaller than the threshold value. Photos 5aýc show three processed images, with each image processed with a different threshold value. As Photo 5c indicates, care must be taken when choosing the threshold value because a great deal of information may be lost if the threshold is too large or small.

a)	b)	c)
Photos 5aýcýHere I used a threshold to create a binary image. aýThreshold set to 15. býThreshold set to 30. cýThreshold set to 45.

Replacing every pixel value in an image with its opposite value (0 becomes 63, 63 becomes 0) produces a negative. Photo 6 shows the negative of the source image. Creating a negative after the image has been stored takes more time than doing it on the fly with an appropriate set of ILUT or OLUT data.

Photo 6ýThis is the negative of the source image. Sometimes an image detail that is not obvious in the normal image is apparent in the negative.

Photos 7aýc illustrate some simple geometric operations. Flipping the image upside down or left-to-right is useful when you cannot control the method of image capture. For example, someone else might give you an image to process, and the image may contain letters and numbers that are backwards. Flipping the image around may help with the processing in this case.

a)	b)	c)
Photos 7aýcýHere simple geometric operations are applied to the source imageý(a) horizontal flip, (b) vertical flip, and (c) zoom.

Zooming in on a region of the image is also an important operation. In Photo 7c, the 128 ý 128 center portion of the source image has been magnified to 256 ý 256.

In general, you may only want to flip or zoom in on a small portion of an image. In this case, additional software must be used to allow you to select the rectangle of pixels that will be processed.

PREVIOUS • NEXT