The intercept method is a convenient, yet powerful method for the nume
rical analysis of fabrics. For geological applications it can be used
when a set of objects (e.g. pores, grains or aggregates of a specific
mineral, lineaments, etc.) with unique characteristics can be identifi
ed in an image. The intercept method analyses boundaries of objects as
a population of lines. How intercepts are definined is therefore impo
rtant in order to establish the significance of the intercept counts.
When the method is applied to digital images, the subdivision of the i
nformation into square pixels must also be considered. We apply a line
ar filter which minimizes the effects of grid anisotropy on the counti
ng of intercepts. Analysis of the resulting rose of intercept counts b
y a Fourier series considerably facilitates the interpretation of the
data. A rose of directions can readily be derived for the phase bounda
ries. If a population of objects can be assumed to have started with a
n originally isotropic orientation and to have been deformed passively
, the Fourier components of the rose of intercepts also give the secti
onal strain. The Fourier components also permit an objective test of t
he validity of describing a fabric as resulting from a homogeneous str
ain on an initially isotropic section. Two computer programs for the c
ounting and analysis of intercepts on digital images are presented and
applied in four examples which illustrate the usefulness of the inter
cept method for magmatic fabric measurement, strain analysis and linea
ment analysis. We also illustrate the importance of Fourier series dec
omposition of rose of intercept counts for fabric interpretations. Int
ercept results are also shown to be compatible with those derived by o
ther methods (eigenvector, SURFOR, inertia tensor, autocorrelation met
hods).