Features and affinities: probing SURF features

The following page documents some probe made with Surf features. Surf stands for Speeded Up Robust Features1. Wikipedia describes the role of the features:
"For any object in an image, interesting points on the object can be extracted to provide a "feature description" of the object. This description, extracted from a training image, can then be used to identify the object when attempting to locate the object in a test image containing many other objects. To perform reliable recognition, it is important that the features extracted from the training image be detectable even under changes in image scale, noise and illumination. Such points usually lie on high-contrast regions of the image, such as object edges."
The Surf features are therefore a sort of DNA or footprint for an image.

The first image shows the opencv2 default ouptut for the SURF algorithm applied to an image of the collection of Guttormsgaard. The circles represent the "zones of influence" of the features

In our first probe, we have tried to isolate the features and look at their visual content. A sample of a few features.

The following images show a panorama of the features. The first one is in black and white as the algorithm operates on black and white versions of the image.

The second is in color. The zones identified in black and white have been replaced by their corresponding zones in the original colored image.

Continue reading

Notes

1. SURF (Speeded Up Robust Features) see Wikipedia entry The SURF algorithm has been patented by the firm Kooaba later sold to Qualcomm. The regulation over software patents in Europe is still a hot debate and the nation states interpret it differently (ie. France rejects it alltogether). The next step of our investigation using feature invariance will include patent-free algorithm like BRISK and ORB

2. OpenCV (Open Source Computer Vision Library: http://opencv.org) is an open-source BSD-licensed library that includes several hundreds of computer vision algorithms.