LIP edge detection

Defined in 1985 by Michel Jourlin, the LIP Model (Logarithmic Image Processing) is a mathematical and physical framework dedicated to image processing. First defined as a framework adapted to images acquired in transmitted light, the interest of using LIP operators in a reflected light model has then been established.
The LIP model is based on the transmittance law. The addition law of the LIP model can be represented as the « addition » of two obstacles placed between a source and a sensor. From this addition law, LIP subtraction and LIP multiplication have been defined.
Since, many applications have been developped around these basic operators : LIP contrasts, LIP interpolation, LIP metrics, …
This plugin proposes an edge detection function based on the LIP additive contrast. On each pixel of the image, a contrast is computed with its neighbours (you can choose between average contrast or maximum contrast).
The advantage of such a contrast lies in its robustness to brightness variation. For instance, the same resulting image will be obtained with input images of a same scene acquired at different exposure levels (in a limited level range). Or, as shown in the illustration, a same texture located in different parts of an image will produce the same contrast values wether it is in a bright or dark area.
For more information, see the following paper.