Auto-Setting of Optimal Exposure for Structured Light 3D Cameras

The quality of 3D point cloud captured by a structured light 3D camera, such as the density and accuracy of the points, depends upon the optimal settings of 3D camera exposure time. The fixed exposure time often set manually, may not be effective under the variation of ambient conditions, e.g., surface reflectance and illumination that incur saturation and/or under-exposure. This is especially the case where 3D cameras are to apply to industrial inspection. In this paper, a novel method for automatically determining the optimal exposure time of a 3D camera based on estimating the pixel intensity-shutter time slope is proposed. The experiments verify that a dense yet accurate 3D point cloud can be captured under varying ambient conditions: up to 40% improvement over the manual setting for the density of captured point cloud.