1.2. Image properties

In order to give a formal definition of an image we first have to answer the question: “Which properties do images typically possess and how can we categorise images based on these properties?”

The first property to discuss is the underlying dimension of the space on which an image is defined. For example in case of a clinical ECG signal, the resulting line plot can be interpreted as a one-dimensional image. The most common type of image, such as a digital photography consisting of pixels, is obviously a two-dimensional image, while a whole-body CT scan of a patient leads to a three-dimensional image comprised of voxels.

Another property takes temporal information into account. Images can either be static, e.g., a photography, or in case of a video sequence by dynamic.

The next property differentiates images by their signal values. In particular, the image range lets us differentiate between monochrome images, in which an object is captured only using the shade of one color (typically gray values), RGB images, which uses the three primary colors red, green and blue to display an object, or hyperspectral imaging as used in remote sensing, in which thousands of channels are used to capture image information.

Finally, we can differentiate between structured and unstructured images. The most common form of pictures consist of rectangular pixels on an image grid, which leads to structured images. On the other hand, modern sensors such as LiDAR can capture the geometry of surfaces via point clouds that lead to unstructured image information.