We derive an algorithm for reconstructing contours of an object from 3
-D cone beam x-ray data. By reducing the amount of the searched-for in
formation, contours, or density jumps instead of the densities themsel
ves, we are able to develop fast algorithms for data incomplete with r
espect to both the movement of the x-ray source and the detector readi
ng. The method is related to local or Lambda tomography. Numerical sim
ulations show the efficiency of the algorithm.