||As product varieties increase and life cycles shorten, the need to reduce product development time becomes more critical to maintain competitiveness in the market. The reduction of product development time, therefore, requires revolutionary improvements rather than gradual changes in technology. Compared to conventional surface measurement techniques, digital moiré technique has many merits: easy implementation; fast full-field measurement; adjacency information between points is readily obtained. Phase unwrapping is a very crucial step of the digital moiré technique to recover the true, depth-encoded phase from its wrapped phase formats not only in the optical measurement but also in MRI (Magnetic Resonance Imaging) or SAR (Synthetic Aperture Radar). If the noise is free, the phase unwrapping problem will become trivial apparently. Conventional phase unwrapping methods unwrap phase values along a fixed path. They take many measures to identify the error points and avoid them. That is inefficient since the starting point may be in the regions with much error points. Also unwrapping all the points in one wrapped phase map is not appropriate whatever they are useful or not in the following steps. For this reason the quality-guided method is proposed to deal with this problem by determining the integration path dynamically. In this method, pixels with higher quality are unwrapped first. The key step of the quality-guided method is to construct a suitable quality map to guide the unwrapping path because there doesn’t exist a quality map that can be effective in all circumstance. A novel phase unwrapping method has been presented and evaluated in this thesis. This method works by first constructing a quality map based on the characteristic of the active triangulation system fiom the wrapped phase data. The quality map is then used to guide the unwrapping. In order to reduce the execution time of phase unwrapping, we tessellate the whole wrapped phase map into blocks. Blocks with higher confidence will be unwrapped first. Experiment shows that this method can unwrap the wrapped phase map derived from the digital moiré system successfully. In data integration stage, only data with higher confidence will be reserved and used to merge together to form a complete 3D model. In a fully automated surface acquisition system, two fundamental problems need to be solved: deciding which areas of the viewing volume need to be scanned and determining how to position the range scanner to sample them. This is known as the Next Best View problem (NBV). In this thesis, we also study this problem preliminarily and propose a simple method to let the system capture the entire object surface automatically using the digital moiré system.