Chapter 13: Color Feature Detection

With contributions by Arnold W. M. Smeulders and Andrew D. Bagdanov

Differential-based features such as edges, corners, and salient points are widely used in a variety of applications such as matching, object recognition, and object tracking. 1Many applications are based on luminance-based features. In this chapter we discuss algorithms for the detection of color features in images. As we will see, this has several advantages over luminance-based features. First of all, we can apply the photometric invariance theory discussed in Chapter 6, which allows us to detect photometric invariant features. Secondly, color plays an important role in attributing saliency to images.

From a mathematical viewpoint the extension from luminance to color signals is an extension from scalar signals to vectorial signals. This change is accompanied by several mathematical obstacles. Straightforward application of existing luminance-based operators on the separate color channels, and subsequent combination of the results, often fails. For example, combining derivatives with a simple addition of separate channels results in cancellation in the case of opposing vectors [234, 235]. This is illustrated in Figure 13.1. For the blue-red and cyan-yellow edge on the right of Figure 13.1a, the vectors in the red and blue channel point in opposite directions and a summation could result in a zero-edge response, while a prominent edge is clearly present. Also, for more complex ...

Get Color in Computer Vision: Fundamentals and Applications now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.