Moire topography is a non-invasive optical technique which may be used for the detection and follow up of spinal deformities such as scoliosis. The moire technique produces interference patterns (moire fringes) which are very good at showing up dizerences in symmetry on the patient's back. The moire' fringes follow lines of constant height and can therefore be considered analogous to contour lines on an ordinance survey map. The moire' technique is already being used as a tool for qualitative diagnosis but widespread use will wait until the task of interpreting the moire patterns quantitatively has been automated. An automated process means extracting the third dimension from the moire' topogram to produce 30 representations from which various deformity indices may be calculated. Image processing of moire' topograms is not trivial due to the di~culty ofproducing distinct good contrast fringes. The contrast problem is addressed with an interactive contrast enhancing algorithm. Ihe approach to extracting 30 information from the fringes is to tackle 'real world'problems head on; fringes arejrst detected using edge detection techniques and assigning heights to the fringes is accomplished with algorithms designed to cope with noisy and broken jinges.