The current snapshot (1.7-SNAPSHOT) of the imagingbook library adds a new feature for drawing anti-aliased graphics directly on the image raster. This is accomplished by the new class ImageGraphics, which works for images of type ByteProcessor, ShortProcessor and ColorProcessor (there is currently no support for FloatProcessor).

standard graphics	anti-aliased graphics

Note that a similar mechanism is implemented in vanilla ImageJ for “flattening” vector overlays. However, this facility does not require the creation of graphic overlays but operates directly on ImageProcessor objects. The new class also adds several convenience methods for drawing operations with floating-point (double) coordinates.

Typical usage (e.g., inside an ImageJ plugin):

import imagingbook.lib.image.ImageGraphics;

ImageProcessor ip = ... ;   // some ByteProcessor, ShortProcessor or ColorProcessor
try (ImageGraphics g = new ImageGraphics(ip)) {
   g.setColor(255);
   g.setLineWidth(1.0);
   g.drawLine(40, 100.5, 250, 101.5);
   g.drawOval(230.6, 165.2, 150, 150);
   ...
}

See the documentation for ImageGraphics for additional details.

There have been various changes and additions to the imagingbook-common library and the associated ImageJ plugins suite:

• Delaunay triangulation: Two implementations of the Delaunay triangulation are now available which share the same interface. The new implementation (Guibas-algorithm) is roughly ten times faster than the previous version (based on Chew‘s algorithm).
• 2D Geometry: Packages delaunay, mappings and fitting in imagingbook.pub.geometry have been revised. A new interface Point has been added to handle various implementations of 2D points across different packages.
• Discrete Fourier and Cosine Transforms:  The  DFT and DCT libraries (packages imagingbook.pub.dft, imagingbook.pub.dct) and associated ImageJ plugins (in imagingbook_plugins_all/Spectral_Techniques). Added fast implementations based on the JTransforms package by Piotr Wendykier. Note that these are efficient mixed-radix implementations of the FFT that are not restricted to image sizes that are powers of two and thus require no image padding!
• ImageJ plugins for book chapters: Maintenance of the plugin collections for all previous book editions has become too tedious. These sub-modules were thus frozen and will not be maintained (i.e., adapted to new releases of the imagingbook-common library) any longer. A complete and continuously updated collection of plugins is contained (as before) in the imagingbook-plugins-all repository. ZIP files with the plugin sets for previous book editions are found in the new plugins-archive.
• Miscellaneous: The Maven dependency setup has been simplified by automatically copying all required runtime-dependencies to ImageJ’s jar folder.

Links:

• https://github.com/imagingbook/imagingbook-common/releases/tag/v1.6
• https://github.com/imagingbook/imagingbook-plugins-all/releases/tag/v1.6

Delaunay triangulation of a set of detected corner points (click on image to enlarge).

Our ImageJ-based camera calibration library has been updated today, source code has been extensively revised and JavaDoc documentation added throughout. Find everything in this GitHub repository:

https://github.com/imagingbook/imagingbook-calibrate

See the associated report for additional details and theoretical material.

Find a new cheat sheet for using ImageJ’s Java API with methods grouped into useful categories here.

ImagingBook’s core library imagingbook-common provides most of the functionality required by the demo ImageJ plugins that supplement each book chapter. It helps to keep the plugins slim, avoids code duplication, and makes the code easily portable to other environments.

For users who want to use imagingbook-common in their Maven projects, the library is now (starting with release 1.5) available as a public Maven artefact. See here for further details.

Modern Java IDEs, such as Eclipse, IntelliJ or NetBeans are excellent tools for developing ImageJ plugins and should be used for projects of any size. However, beginners sometimes find it difficult to set up their IDE to work smoothly with ImageJ’s file structure. We have created a small repository on GitHub that provides a simple setup to get started with ImageJ plugin development in Eclipse or IntelliJ:

See https://github.com/imagingbook/imagej1-plugins-ide-setup

Mistakes happen and some even remain undetected for quite some time. Many careful readers have helped us to correct smaller and greater errors. This time our special thanks go to Jan Sellner for thoroughly checking through Chapter 17 (Anisotropic Diffusion Filters) of the German 3rd edition. The updated errata pages can be found here.

Today, the imagingbook source code repository was officially moved to GitHub for various technical reasons. Bitbucket has been a great host, providing free private repositories by default – thanks for this and many other great features! While we’ll continue to use Bitbucket for other projects, the main motive for this switch was the easier coupling to the static web pages we use for creating JavaDoc output. We also added a new repository for release (JAR) files that is under GIT version control.

Here is a listing of all relevant repositories:

• Source code: https://github.com/imagingbook/imagingbook-public
• JavaDoc pages: https://imagingbook.github.io/imagingbook-doc
• Release files: https://github.com/imagingbook/imagingbook-jars

Related repositories (projects using the imagingbook library):

• Camera calibration: https://github.com/imagingbook/imagingbook-calibrate
• Viola-Jones Face Detection: https://github.com/imagingbook/imagingbook-violajones

Please visit our main site imagingbook.com for additional details.