2. Geometry

A vision system is a chain for converting light into an image. Several elements make up this chain: the optics, the sensor and the conversion electronics.

In order to exploit the information extracted from images, it is necessary to model the entire acquisition process. In our case, this means writing the equations that govern image formation from a geometric point of view. It introduces the parameters that characterize the system. These parameters then need to be estimated, so that measurement, quality control and 3D reconstruction can be carried out.

To estimate the parameters, information must be extracted from the acquired images. Depending on requirements, this information, commonly referred to as "primitives", can be points, contours, straight lines, circles and so on. In this article, we'll confine ourselves to the use of points as...