The registration of point clouds in urban environments faces problems such as dynamic vehicles and pedestrians, changeable road environments, and GPS inaccuracies. The state-of-the-art methodologies have usually combined the dynamic object tracking and/or static feature extraction data into a point cloud towards the solution of these problems. However, there is the occurrence of minor initial position errors due to these methodologies. In this paper, the authors propose a fast and robust registration method that exhibits no need for the detection of any dynamic and/or static objects. This proposed methodology may be able to adapt to higher initial errors. The initial steps of this methodology involved the optimization of the object segmentation under the application of a series of constraints. Based on this algorithm, a novel multi-layer nested RANSAC algorithmic framework is proposed to iteratively update the registration results. The robustness and efficiency of this algorithm is demonstrated on several high dynamic scenes of both short and long time intervals with varying initial offsets. A LiDAR odometry experiment was performed on the KITTI data set and our extracted urban data-set with a high dynamic urban road, and the average of the horizontal position errors was compared to the distance traveled that resulted in 0.45% and 0.55% respectively.

Since wear particles usually contain valuable information about lubricant and wear condition of lubricated components, wear debris analysis is widely used to monitor machine/component wear. Current literature reports about the importance and necessity of 3-dimensional (3D) surface characterization of wear particles as opposed to 2D characterization. It may be borne in mind that traditional 3D analysis has been a synonym only for local surface morphologies for stationary particles. This results in large errors when irregular particles are under observation. A new 3D imaging methodology is proposed to reconstruct multi-view 3D surfaces for a moving particle whereby the target particle is passed through a micro-channel with rolling motion, and recorded via a digital microscope. The multiple moving particles are tracked based on Kalman filter to obtain different-view images. Sparse and dense reconstruction of the particle surfaces are enabled by these images using Structure from Motion (SfM) and Shape from Shading (SfS) methodologies, respectively. The method was verified using real particles and evaluated by the results of Laser Scanning Confocal Microscopy. The proposed method is able to reconstruct wear particles under various perspectives, and the errors from the characterization parameters of the Areal are less than 20%.

Wear debris analysis (WDA) enables the provision of essential information towards the monitoring of machine fault diagnosis and the analysis of wear mechanism. However, this experience-based technology has not yet been automated for the identification of similar particle types due to the small number of samples and highly dispersed features. To address this problem, a knowledge-guided convolutional neural network model is developed to focus on two representative severe wear particles: fatigue and severe sliding particles that have highly similar contours but weakly discriminative surfaces. The height images of particle surfaces are adopted as the initial objective. Characterized by typical particle features, the empirical WDA knowledge is represented into the feature-marked images, and further automatically learned by a U-Net-based knowledge extraction network. By weighting with the U-Net output, a knowledge-guided particle classification network is constructed to identify similar particles under a small number of samples. With this methodology, the empirical WDA knowledge is transferred to guide the classification network for locating the discriminative features in particle height images. Thirty sets of fatigue and severe sliding particles are acquired from wear tests as the training and testing samples. For verification, the network kernel is visualized to trace the particle feature propagation in the classification. Experimental results reveal that the proposed method can accurately identify fault particles that are acquired from wear tests.