||The rigorous requirements for reinforcing bar (rebar) surface geometry were raised by local government recently in line with the newly released British Standards. To some extent, such requirements guarantee the safety of constructions built with reinforced concrete. The new Standard defined six key parameters to determine the quality of rebar. However, no equipment for measuring all six parameters is currently commercially available to substitute manual measurement. In this thesis, a system is proposed to accomplish measurement of rebar surface geometry automatically and efficiently. To deal with the measurement, it is necessary to divide the whole process into two parts. One part is to generate a 3D point cloud of the rebar surface. The other is to determine the parameter values of the surface from the point data. At first, the method of reconstruction via stereo vision is adopted by excluding any other possible solution in this case, since only the sampling resolution of CCDs could be possible to fulfill the requirements and efficiently realize measurement of multiple points. Two views of the same part of the rebar are used to reconstruct 3D points by triangulation. Then, after eliminating noise within the point data, six parameters are efficiently determined by robust algorithms using projections on multiple cross section planes in space. Now it is feasible to obtain the six key parameters efficiently and effectively in a few minutes. These calibration methods guarantee the accuracy of the parameter values. Results from several repeated tests show the robustness of the algorithms. Although the system is still in the prototype stage, the problem for measuring rebar surface is successfully solved. Based on similar principles, it is possible to apply such methodology to reconstruct and measure any other object in different scales.