This lesson introduces basic notions related to image processing, i.e. image representation and analysis. Various aspects of image processing are represented in the bloc diagram of figure 1, which will constitute our guiding principle.
More precisely, the lesson will be decomposed in five parts corresponding to the first two blocks of the previous diagram. The third and last block (“utilization”) includes high-level processing such as edge detection and will not be treated in this lesson.
Concerning the acquisition part, and before any further consideration, we will present the diverse representations of an image considered as a bi-dimensional signal. A few examples of basic bi-dimensional signals and their notations will be described. Since image representation in the frequency space is of importance both for low-level and for more sophisticated processing, we will describe the properties of two-dimensional Fourier Transform (FT) in the second part of the lesson. In a third part, the FT properties will be used to introduce some important elements of image filtering that will be present in almost every low-level processing techniques, also called image preprocessing. The problem of image sampling and quantization will be presented in the fourth part. Finally, digital image filtering will be described in the last part of the lesson.