Feature enhancement for a defocusing structured-light 3-D scanning system

Abstract

Structured-light systems consisting of a camera and projector are powerful and cost-effective tools for three-dimensional (3-D) shape measurements. However, most commercial projectors are unable to generate distinct patterns due to defocusing and shallow focusing issues. We propose a hybrid method for enhancing the calibration and scanning features of the defocusing structured-light 3-D scanning system. Instead of using conventional sequential binary patterns, we replace the highest-level binary pattern by a high-order sinusoidal pattern. In our proposed system, a pan-tilt stage carrying a checkerboard is used to assist the simultaneous calibration of the camera and projector. Initially, the camera is calibrated to obtain the extrinsic positions of the stage. In addition, we utilize the multiplication of vertical and horizontal stripe patterns to enhance the corresponding features between the camera and projector. The projector is then calibrated using the extrinsic features determined from the calibrated camera. The experimental results show that the use of the high-order sinusoidal pattern significantly improves reprojection error. Our proposed method can easily be incorporated in the defocusing projector for scanning various types of objects.

Industrial Robotic Arm in Machining Process Aimed to 3D Objects Reconstruction

“…In 2018, Oliveira used an embedded system to automate the acquisition of images for 3D acquisition using photogrammetry [9]. On the other hand, Liu and Lin tested a high-order of sinusoidal patterns in addition to conventional sequential binary patterns to improve scanning and reduce errors in the 3D acquisition by Structured-light [10]. In addition to these projects, there were projects in 2017 dealing with the integration of 3D acquisition with a 3D printer [11] and CNC machining, when Santos used a CAD model properly acquired and treated, in a CAM software to obtain the G-code, to milling the part on a CNC machine [12].…”