Raster Image - Feature (Image Properties)

Global

For the whole image

Local

For a set of pixel.

Local features provide a limited set of well localized and individually identifiable anchor points. Their location must be determined accurately and in a stable manner over time. They can be used as a robust image representation, that allows to recognize objects or scenes without the need for segmentation. Features need to be described, such that they can be identified and matched.

• Local features are covariant.
• The descriptors derived from them, on the other hand, are usually invariant, due to a normalization step.

Example:

• points (interest point)
• region
• edge segment
• small image patches
• corner,
• blob

Usage

• edges detected in aerial images often correspond to roads;
• blob detection can be used to identify impurities in some inspection task;
• matching or tracking applications,
• camera calibration or 3D reconstruction.
• image alignment or mosaicing
• object recognition
• scene classification,
• texture analysis,

Detection

• To localize features in images, a local neighborhood of pixels needs to be analyzed
• spatial extent: the local neighborhood of pixels