Book Review: Digital Color Image Processing by Andreas Koschan and Mongi Abidi

Everywhere you look the significance of color information in digital processing is becoming more important than ever. The manner that color is addressed in many systems is via monochromatic based techniques; that is breaking things down into channels and manipulating them using each of the color components.

What has not been addressed in many textbooks is making the transition from scalar to vector-valued image functions. This is where Digital Color Image Processing comes in. It was written to present a detailed introduction to this topic

Digital Color Image Processing is a textbook organized in regard to advanced techniques for three dimensional scene analysis. It is contained in 376 pages, 13 chapters, and is fully indexed. It breaks down into four fundamental sections: fundamentals, processing, scene analysis, and case studies.

Chapters 1-4 begin with an introduction to the book and to the terminology used in color image processing. With regard to grey-level images the terms are common, but when trying to transfer this to vector-valued images, a common language does not yet exist.

From there, an introduction to human color vision is presented and the color blindness of a selection of the population is also addressed. Then the topic of color spaces used in color imaging is presented. Finally, the technical requirements for color image processing are considered. This includes camera, filter, illuminates, and charts, as well as techniques of photometric and colorimetric calibration needed for treatment of color images.

Chapters 5-8 focuses in on noise suppression and contrast enhancement in color images. Then the process of extraction of edges in an image is discussed. There are various procedures that are addressed and a comparison of the results of one monochromatic based and two vector-valued color edge operators are also given.

Then the techniques of using color information for image segmentation are analyzed. Here, four classes of segmentation are introduced and discussed. Finally, an overview of the techniques for highlight analysis and a new method for minimizing inter-reflections in real color images is presented, as well as alternate procedures for achieving color consistency is examined.

Chapters 9-10 describe the use of color information for static stereo analysis. By viewing objects from different locations, one can better determine their positions in a three dimensional area. Investigations for edge-based as well as area-based color stereo techniques are discussed. You will see how stereo matching results can be significantly improved by projecting color-coded light patterns onto the object.

Here also, the effects of motion or photometric changes due to lighting are considered, as is the inclusion of color information into dynamic and photometric stereo analysis.

Chapters 11-13 address case studies of color use in an automated video tracking and location system that is under development at the University of Tennessee’s Imaging Robotics, and Intelligent Systems (IRIS) Laboratory. Covered here is the acquisition and analysis of multi-spectral images and its use in face recognition as well as the application of color coding in x-ray imaging in luggage scans to identify threats during airport screenings.

In addition to introducing you to important new technologies in the field, Digital Color Image Processing also covers topics such as techniques for improving three-dimensional reconstruction, three-dimensional computer vision, and video surveillance. Digital Color Image Processing is well written. It will take the reader beyond the standard of image processing and should prove to be a complement to existing textbooks in its field.

Digital Color Image Processing is a technical book aimed at the student, as well as the professional, who is trying to get a handle on vector-value image functions. This book is recommended for first and second-year graduate students in electrical and computer engineering and computer science courses, as well as for researchers with basic knowledge in image processing who wish to extend their knowledge in the area of color image processing.