The quality of reconstructed surfaces from measured data is related to many factors, such as quality of original data and the parameterization results of them, surface fitting strategy, surface post-processing, and so on. This dissertation, in a certain extent, has studied the relative techniques of triangular mesh based high quality surface reconstruction and surface interrogation. Main contents and research achievements are as follows:Mesh smoothing and regularization techniques has been studied. Based on the new formula for the calculation of mesh vertex normal vector and discrete curvature, a novel approach for triangular mesh smoothing and regularization are presented. The main advantage of the approache is the lower distortion and feature-preserving.The problem of parameterization for triangular mesh is studied. Based on improvements to Sheffer's ABF method, a novel parameterization method is proposed. It solved the ABF's problem of parameter domain border, such as intersection and non-convex.The problem of triangular mesh based C2- continuous piecewise bi-cubic B-spline surface reconstruction has been studied. A novel surface reconstruction algorithm is put forward. It can join the reconstructed surface patches along their co-edges and at their confluent point C2-continuosly and seamlessly, so can guarantee the whole surface is C2- continuous everywhere.The problem of surface interrogation and surface fairing has been studied. Firstly, by improving the existing light-cylinder model based highlight-line method for surface interrogation, a new method, named light-strip model based highlight-line method, is proposed. It overcame the disadvantages of the existing one. Secondly, an improved approach for fairing of bi-cubic B-spline surface is presented. This approach adoptsrevision of point selectionstrategy, holds local character, and is efficient.Experimental results are presented to demonstrate the effectiveness and efficiency of the proposed approaches.