Depth measurement is an important problem in computer vision field. Firstly, a new grid pattern was designed and a sequence coding and decoding algorithm for the pattern was proposed. Secondly, a linear fitting algorithm of system parameters for constructing the linear relationship between the object depth and the pixel shift was proposed, and the lens distortion was rectified. Thirdly, depth information of object was obtained based on this linear relationship. Moreover, 3D reconstruction was implemented based on Delaunay triangulation algorithm. Finally, the regularity of the error curves was utilized to correct the system errors and the measurement accuracy was improved. The experimental results show that the accuracy of depth measurement is related to the step length of moving object.