基于相移法的多目标运动物体三维重构

相移法可实现静态物体三维形貌的高精度重构,对于运动物体形貌重构则误差较大。其根本原因为相移法需要多个条纹图进行物体重构,而传统相移法理论没有包含物体的运动信息,无法描述物体运动对相位的影响。导致当物体在条纹图间发生运动时测量误差较大。针对以上问题,提出了一种利用物体运动信息对多个二维运动物体进行三维重构的新方法。不同的被测物体可具有不同的运动轨迹。首先,对多个被测物体进行识别并确定目标区域;然后,采用KCF算法对物体进行跟踪并使用SIFT算法提取物体运动前后的特征点,分别估计描述物体运动的旋转平移矩阵。将运动信息带入条纹描述方程中,获得包含运动信息的三维重构模型,最终采用最小二乘法提取正确的相位值。结果证明:该方法能有效地减少由物体运动引起的测量误差,扩展了三维测量的应用范围,具有较高的工业应用价值。

3D reconstruction of multi-target moving objects based on phase-shifting method

Phase shifting profilometry (PSP) can reach high accuracy for the 3D shape measurement of static object. However, errors will be introduced when moving object was reconstructed. The fundamental reason was PSP required multiple fringe patterns to reconstruct the object and the traditional PSP did not contain movement information. Aiming at the above, a new method for the 3D reconstruction multiple 2D moving objects was proposed. Different objects can have different movement. Firstly, the multiple objects were identified and the areas of interests are defined. Then, the KCF algorithm was used to track the moving object and SIFT algorithm was used to retrieve the feature points of the object before movement and after movement. The rotation matrix and translation vector describing the movement was then obtained. With the help of the reconstruction model with movement information, the least-square algorithm was employed to retrieve the correct phase value. The results show that the proposed method can reduce the errors introduced by the movement and has the potential to be applied in industrial field.