Abstract:The iterative closest point (ICP) algorithm is one of the most widely used methods for point sets' registration. However, ICP is very sensitive to the selection of initial points and is easy to fall into local optimum. To address this problem, many techniques have been developed. In this study, a two-step registration method is proposed for two 3-D point sets' registration, which is achieved by a combination of rough and fine registrations. Specifically, a multidiscriminant parameter feature (MDPF) extraction approach is developed and embedded into the rough registration stage in order to find new corresponding point pairs for the fine registration. Three geometric features are chosen after experimental investigation for key points selection. By using the threshold discriminant condition to determine the key points and the distance constraint to eliminate the wrong point pairs, the feature points can be extracted, and the final transform parameters can be derived based on these feature points. In order to improve the computational efficiency for fine registration, the center of gravity is created to find the closest point in solving the transformation matrix, which is especially beneficial for registering complex surfaces. Experimental results show that the proposed method outperforms the traditional ICP approach and some typical existing improved algorithms in terms of the root-mean-square error (RMSE), the total number of feature points, and the execution time. In particular, the performance is improved 40% in terms of the RMSE and 50% in terms of the execution time in comparison with ICP on some benchmark data sets. Experiments also demonstrate that reliable reconstruction results can be obtained for both real outdoor and indoor environments.

Precision Registration of Corresponding Uncode Points in Large-Sized 3D Surface Measurement Using Three Images

A New Method for Registration of 3d Point Sets with Low Overlapping Ratios

Large 3D free surface measurement using a mobile coded light-based stereo vision system

Multi-camera Layout Design on Large Scope Visual Inspection and Its Data Registration

Novel 3-D Coordinate Measuring System Based On Binocular Coded Points Matching

3D Datum Mosaic Method of Multi-View and Large Scale Based on Optical Coding

An Automatic Registration Approach to Laser Point Sets Based on Multidiscriminant Parameter Extraction

Study on High-Precision Inspection Multi-Vision Cameras Space Layout

Novel method of measurement for large-space 3D coordinate by binocular code matching

All-sided 3D Non-connecting Detection to Objects with Large Surface Areas

Mobile 3D vision measurement system for large-scale free form surface

High-accuracy Point Cloud Registration for 3D Shape Measurement Based on Double Constrained Intersurface Mutual Projections

3D Measurement of Large and Complex Parts Based on Phase Matching and Global Marker Points Registration

Optical and Contact Probe for Industrial Measurement of 3-D Points

A new 3D high accuracy optical coordinates measuring technique based on an infrared target and binocular stereo vision

3D Measurement of Large-scale Object Using Independent Sensors

Optical-probe-imaging-based stereo cameras 3D coordinate measuring system

Global Data Registration Technology Based on Dynamic Coded Points.

New Measuring Method of Large Size Measurement Based on Close Range Photogrammetry and 3D Optical Measurement

Study of Stereo Vision Coordinate Measurement System Based on Optical Probe

Precision Evaluation of Three-dimensional Feature Points Measurement by Binocular Vision