散斑投影轮廓术通过投影单幅随机散斑图案编码场景的深度信息，利用散斑匹配技术建立立体图像间的全局对应关系，从而实现单帧3D重建。但由于被测物体表面的复杂反射特性和双相机间存在的视角差异，投影单幅散斑图案无法为整个测量空间中每个像素编码全局唯一的特征，由此带来的误匹配问题导致测量精度较低，难以满足一些工业场景的高精度测量需求。提出一种基于垂直腔面发射激光器（VCSEL）投影阵列的散斑结构光三维成像技术及其传感器设计方法，所研制的三维结构光传感器集成了3个小型化散斑投影模组，投影一组空间位置不同的散斑图案，对被测场景的深度信息进行高效时空编码。提出一种由粗到精的时空散斑相关算法，以提升测量精度，重建复杂物体的精细轮廓。通过精度分析、三维模型扫描、小目标金属零件检测、复杂场景测量等实验证明，所提三维结构光传感器实现了远距离、大视场的高精度三维测量，可潜在应用于零件分拣、机器人码垛等工业场景。

Speckle projection profilometry (SPP) encodes depth information of a scene by projecting a single random speckle pattern and establishes the global correspondences between stereo images using speckle matching technology, thereby achieving a single-shot three-dimensional (3D) reconstruction. However, it is still difficult to encode a globally unique feature for every pixel in an entire measurement space by projecting only a speckle pattern due to the measured object surface with complex reflection characteristics and the perspective differences between stereo cameras. The resulting mismatching problem leads to the low measurement accuracy of SPP, making it difficult to meet the high-precision measurement requirements of some industrial scenarios. In this study, we propose a speckle structured-light-based 3D imaging technique and its sensor design method using a vertical-cavity surface-emitting laser (VCSEL) projection array. The proposed 3D structured-light sensor integrates three miniaturized speckle projection modules to project a set of speckle patterns with different spatial positions, spatially and temporally encoding the depth information of the measured scene efficiently. In addition, a coarse-to-fine spatial-temporal speckle correlation algorithm is proposed to improve the measurement accuracy and reconstruct the fine shapes of complex objects. Experiments, including precision analysis, 3D model scanning, detection of small target metal parts, and measurement of complex scenes, indicate that the proposed 3D structured-light sensor can realize high-precision 3D measurements at a long distance over a large field of view, which can further be applied to industrial scenarios such as part sorting and robot stacking.