|| ||           || Doxygen's  

---

Image definition and generic operations

Images and semantics

The use of images include:

• intensity images: the pixel value indicates a monochrome light intensity
• color images: the pixel value represents a color (RGB, HSI, etc.)
• X-ray, ultrasound, or electron microscope images: a pixel value depends on object density or another physical phenomena
• satellite images: the pixel value represents a recording of up to 7 spectral bands
• range images: the pixel value indicates a distance
• characteristic images: the pixel value indicates whether the pixel is element of a set
• flow fields: the pixel value represents a motion vector
• complex images: FFT domain

As stated in the introduction, we strive for a separation between semantics, representation, and implementation of objects. The representation is to provide a ``neutral'' view of a concept, a more or less mathematical view on the functionality needed in dealing with instances of the concept. From a software design point of view, representation and associated functionality is more important than semantics because it provides most opportunities for reusability. After all, you could say that semantics is nothing more than a label assigned to a representation.

Images in Horus

In Horus images are represented by the class HxImageRep. The actual image data is stored in HxImageData. The class Image with semantics is used to associate semantics with the pixel values, i.e. what does the numerical value of a pixel actually mean. The current implementation does not provide a class for semantics.

Image definition

Generic image operations

---

Go to the next section or go to images or return to the index.

---