for any quantification I suggest firstly to work "nose-on-rock" and subdivide your outcrop into different squares, framed with yardsticks, within which you can describe and quantify the planar architectural elements, then you can take different photographs for further investigation, proceeding from the large-scale to the small-scale mapping (in this mapping of a cross section). For further reading I suggest the textbook on "Geomorphology" issued by Summerfield M.A. for classification. This appraoach can be taken for engineering geology, applied sedimentology and hydrogeology. If you are going to disclose your aims and give a bit more detailed insight into the local geology I can likely fine-tune my answer, which at the moment is kept a bit more general.
I had never to face this issue so far but, to answer your request, I agree with Harald that you should first try to collect as many detailed information from the outcrop itself as possible (percentage of empty vs clay-filled fractures, azimut and dip wrt slope gradient vs depth of observation, root control on fracture widening, etc). Then, I would suggest you take digital photographs of the outcrop at different representative places and proceed with image analysis using ad hoc sofwares and methods (edge enhancing, fracture mapping and statistical analysis: fracture length and width distributions, spacing, continuity and networking, density distribution, etc - have, e.g., a look at: http://www.biotechniques.com/multimedia/archive/00004/BTN_A_05396ST04_O_4994a.pdf).
you need to study of rock classification ,It is different for volcanic rocks,metamorphic rock and sedimentary racks. fractures are tension ,sheer a combination.and bedding effect in landslides.
it depends on the scale of the work and fractures, we begin by mapping the frcaturation by following a statistical study, direction length and also their frequency per square meter or square km,;;;; (depending on the scale) and the fratures type, diaclase, cleavage plan or fault. for more details you can see this work: Reda Samy Zazoun (2008): The Fadnoun area, Tassili-n-Azdjer, Algeria: Fracture network geometry analysisJournal of African Earth Sciences 50 (2008) 273–285
Detailing regional fracture/lineament characteristics is a first step in a good research analysis. Remote sensing datasets such as SRTM, NED, ASTER can be processed as directional hillshades to expose fractures that can be mapped and characterized in stereo net programs. Knowing the tectonic and structural history of a region makes determination on a local scale much easier.
Thank you for your advice. The characterization is not done on a small scale and not on aerial photographs but at different outcrop according to lithology (phylittes, schist).
Alain thank you for your valuable advice, you have noticed that the outcrop is in the Ardennes. The characterization will be made to constrain a hydrological model at the level of porosity, water storage. The soil is characterized but the basal part of the periglacial layer and the beginning of the open cleavage schist is more difficult
If you are interested in measuring the hydraulic properties of individual fractures or smaller fracture networks on an outcrop, I suggest you get a hold of a hand held permeameter. The one I know best is the TinyPerm instrument designed by and built at New England Reasearch, Inc. (NER) in White River Junction Vermont. Check it out. Nothing beats direct measurement provided you know the boundaries of its application.