Voxel-based Path Planning for 3D Scanning of Mechanical Parts

Abstract

The paper deals with an original approach to scan path planning that applies for any type of sensors. The approach relies on the representation of the part surface as a voxel map. The size of each voxel is defined according to the sensor FOV. To each voxel, a unique point of view is associated in function of visibility and quality criteria. Whatever the sensor, the method provides a set of admissible points of view to ensure the surface digitizing with a given quality.

Simulation of Laser-Sensor Digitizing for On-Machine Part Inspection

“…Yang and Ciarallo use a genetic algorithm to obtain a set of viewing domains and a list of observable entities for which the errors are within an acceptable tolerance [8]. The approach developed in [9] relies on the representation of the part surface as a voxel map, for which the size of each voxel is defined according to the sensor field of view (fov). To each voxel, a unique point of view is associated in function of visibility and quality criteria leading to a set of admissible viewpoints to ensure the surface digitizing with a given quality.…”

Scanner path planning with the control of overlap for part inspection with an industrial robot

“…Yang and Ciarallo use a genetic algorithm to obtain a set of viewing domains and a list of observable entities for which the errors are within an admissible tolerance [10]. The approach developed by Lartigue et al [11] relies on the representation of the part surface as a voxel map, for which the size of each voxel is defined according to the size of the scanner FOV. To each voxel, a unique point of view is associated according to visibility and quality criteria.…”

“…The FOV is the part of the laser beam which is visible by the scanner camera ( [11]). Considering the laser-scanner mounted on a CMM, Bernard and Véron [2] introduce three levels of visibility (local, global and real) to generate a sensor trajectory welladapted to the control of complex parts.…”

Conversion of 3D scanned point cloud into a voxel-based representation for crankshaft mass balancing

“…The scanning methodology is defined by a set of points of view determined by the Voxel2scan method [21]. Each point of view is defined by a position and an orientation in the space in which the scanner has to be placed in order to digitize the maximum surface of the object and to minimize the measurement time.…”

“…In order to obtain the same conditions for scanning, the Voxel2scan path planning method [21] is applied to the selected digitizing system. Voxel2scan algorithm is an original approach to generate a path planning for scanning objects with any geometry and with any type of 3D scanning optical sensors.…”