Flexible High-Resolution Continuous 3-D Scanning for Large-Scale Industrial Components

Abstract

Generating a large-scale high-density point cloud is crucial for three-dimensional shape measurements. Both area-scan and line-scan camera systems have advantages and limitations for large-scale structured light measurement. Consequently, we propose a method that achieves complementary advantages of both the camera systems through flexible continuous scanning with high resolution and frequency. For point cloud generation, efficient scanning in continuous motion with high resolution, large field of view, and high frequency is realized using line-scan cameras. A flexible registration strategy without mechanical positioning is proposed for point cloud registration after referring to the registration methods in area-scan camera systems. To realize the above functions, system structure and calibration are designed. Techniques for point cloud generation with high dynamic performance and interference resistance are introduced. The point cloud registration strategy includes methods for high-frequency pose acquisition and processing. Moreover, an optimization method is designed for the point cloud after registration, and the texture is attached. The proposed system has excellent prospects for large-scale industrial component measurements.