Featured White Papers

• What Developers Want: The End of Application Redeploys

Sponsored Links

• Career opportunities at Harman- Careers That’ll Rock Your Socks Off.

  Learn More!

• Developer-friendly Java PDF libraries by Qoppa Software

  Live Demos

• Stop blaming your app! Diagnose Java performance issues.

  Free tool >

• UrbanCode: Go From Continuous Integration to
  Continuous Delivery

  Learn More!

Sponsored Links

Optimize with a SATA RAID Storage Solution
Range of capacities as low as $1250 per TB. Ideal if you currently rely on servers/disks/JBODs

Image processing with Java 2D

Sophisticated image processing support comes to Java in the 2D API

By Bill Day and Jonathan Knudsen, JavaWorld.com, 09/01/98

If you did any image processing work in JDK 1.0 or 1.1, you probably remember that it was a little obtuse. The old model of image data producers and consumers is unwieldy for image processing. Before JDK 1.2, image processing involved MemoryImageSources, PixelGrabbers, and other such arcana. Java 2D, however, provides a cleaner, easier to use model.

This month, we'll examine the algorithms behind several important image-processing operations (ops) and show you how they can be implemented using Java 2D. We'll also show you how these ops are used to affect image appearance.

Because image processing is a genuinely useful standalone application of Java 2D, we've built this month's example, ImageDicer, to be as reusable as possible for your own applications. This single example demonstrates all of the image-processing techniques we'll cover in this month's column.

Note that shortly before this article went to publication, Sun released the Java 1.2 Beta 4 development kit. Beta 4 seems to give better performance for our example image processing operations, but it also adds some new bugs involving bounds checking of ConvolveOps. These problems affect the edge detection and sharpening examples we use in our discussion.

We think these examples are valuable, so rather than omit them altogether, we compromised: to ensure it runs, the example code reflects the Beta 4 changes, but we've retained the figures from the 1.2 Beta 3 execution so you can see the operations working correctly.

Hopefully, Sun will address these bugs before the final Java 1.2 release.

Image processing isn't rocket science

Image processing doesn't have to be difficult. In fact, the fundamental concepts are really quite simple. An image, after all, is just a rectangle of colored pixels. Processing an image is simply a matter of calculating a new color for each pixel. The new color of each pixel can be based on the existing pixel color, the color of surrounding pixels, other parameters, or a combination of these elements.

The 2D API introduces a straightforward image processing model to help developers manipulate these image pixels. This model is based on the java.awt.image.BufferedImage class, and image processing operations like convolution and thresholding are represented by implementations of the java.awt.image.BufferedImageOp interface.

The implementation of these ops is relatively straightforward. Suppose, for example, that you already have the source image as a BufferedImage called source. Performing the operation illustrated in the figure above would take only a few lines of code: