Many well studied works for accurate 3-D shape measurement which based on structured light technique have been presented in last decades. Among them, one common way is utilizing the binocular stereovision like model to get 3-D information. In this model, projector is always treated like a camera, thus making the projector-camera based system unified with a well-established traditional binocular stereovision system. After calibrating projector and camera, 3-D shape information is obtained by performing conventional triangulation algorithm. However, as we all know that, there is a short baseline problem which highly limits the measurement accuracy in a short baseline arrangement stereovision system. The other widely used method is fringe projection profilometry (FPP). In this work, we present a new and different projecting-imaging model for 3-D measurement error analysis. We first explore the exact mathematical model that exists between the height of an object's surface, the phase distribution map and the parameters of the setup. Then, some practical considerations are taken into account to improve the measurement accuracy. Finally, we study the problem how uncertainty of the parameters in the proposed model affects final measurement accuracy, especially focus on the length of baseline analysis between the projector and the camera. According to the parameters analysis, some experimental results are presented to demonstrate its validity.