Teng Xu

Bio: Teng Xu is an academic researcher from Beihang University. The author has contributed to research in topics: Point cloud & Inertial navigation system. The author has an hindex of 4, co-authored 10 publications receiving 57 citations.

Detection of Water Leakage in Underground Tunnels Using Corrected Intensity Data and 3D Point Cloud of Terrestrial Laser Scanning

Abstract: Detection of water leakage is one of the most important regular tasks for underground tunnel inspection. In this paper, a new method is proposed for the water leakage detection in underground tunnels by using the corrected intensity data and 3-D point cloud of a terrestrial laser scanning (TLS) sensor. In the proposed method, the distance effect on the TLS intensity is first corrected based on a piecewise linear interpolation by using a reference target. Then, the distance-corrected intensity data are used to determine the surface roughness parameter that is specially considered to correct the incident angle effect. After corrections of distance and incident angle effects, the corrected intensity data are used to detect the water leakage regions in the underground tunnel. Finally, the appendages on the tunnel wall are removed by using the 3-D point cloud data to eliminate their influence on water leakage detection. To validate the feasibility of the proposed method, a case study in an underground tunnel in Beijing, China, was conducted by using a TLS sensor. Experimental results have demonstrated that the water leakage regions in the underground tunnel can be well extracted by using the corrected intensity data and 3-D point cloud, especially when the surface roughness is considered.

Terrestrial Laser Scanning Intensity Correction by Piecewise Fitting and Overlap-Driven Adjustment

Abstract: Terrestrial laser scanning sensors deliver not only three-dimensional geometric information of the scanned objects but also the intensity data of returned laser pulse. Recent studies have demonstrated potential applications of intensity data from Terrestrial Laser Scanning (TLS). However, the distance and incident angle effects distort the TLS raw intensity data. To overcome the distortions, a new intensity correction method by combining the piecewise fitting and overlap-driven adjustment approaches was proposed in this study. The distance effect is eliminated by the piecewise fitting approach. The incident angle effect is eliminated by overlap-driven adjustment using the Oren–Nayar model that employs the surface roughness parameter of the scanned object. The surface roughness parameter at a certain point in an overlapped region of the multi-station scans is estimated by using the raw intensity data from two different stations at the point rather than estimated by averaging the surface roughness at other positions for each kind of object, which eliminates the estimation deviation. Experimental results obtained by using a TLS sensor (Riegl VZ-400i) demonstrate that the proposed method is valid and the deviations of the retrieved reflectance values from those measured by a spectrometer are all less than 3%.

A Fuzzy PID Controller-Based Two-Axis Compensation Device for Airborne Laser Scanning

Abstract: In this paper, a two-axis compensation device was designed to eliminate the impact of the attitude deviations of the moving platform on the laser point cloud. In order to improve the dynamic performance of the compensation device, the fuzzy proportional-integral differential (PID) controller was employed for the compensation device. Then, a semi-physical simulation experiment was carried out to quantitatively evaluate the dynamic performance of the compensation device using a three-axis turntable and a high-accuracy inertial measurement unit (IMU), which was used to measure the attitude deviations of the three-axis turntable in real time. As a result, the roll and pitch deviations of turntable measured by IMU can be compensated using the two orthogonal frameworks of the compensation device controlled by fuzzy PID controller. The experimental results showed that the fuzzy PID controller compensated the attitude deviations more accurately than the traditional PID controller and the laser pointing direction could be remained within 0.2° using the compensation device.

GPS-aided method for platform attitude determination based on target images.

Abstract: Attitude measurement error is one of the main factors that deteriorates the imaging accuracy of laser scanning. In view of the fact that the inertial navigation system (INS) with high accuracy is very costly, a low-cost but effective GPS-aided method based on the target images is proposed to measure the platform attitude angles in this paper. Based on the relationship between the attitude change of the platform and the displacement of two adjacent images, the attitude change can be derived by the proposed method. To quantitatively evaluate the accuracy of the platform attitude angles measured by the proposed method, an outdoor experiment was carried out in comparison with the GPS/INS method. The preliminary results demonstrated that the measurement accuracy using the proposed method was better than 0.05° (RMSE).

LiDAR Ranging System Based on Automatic Gain Control and Timing Discriminators

Abstract: A LiDAR ranging system has been developed based on automatic gain control and timing discriminators in this paper. Timing discriminators in the time-pickoff circuit, which employ constant fraction discriminator, were designed to reduce the walk error caused by the variation of pulse amplitude. Automatic gain control was used to concentrate the input pulse amplitude of timing discriminator in a limited range, which can meet the demand of timing discriminators and reduce the timing jitter error of timing discriminators. The time-pickoff circuit using automatic gain control (AGC) and the timing discriminators can improve the precision of LiDAR ranging system. The performance of the designed circuits of AGC and timing discriminators has been tested through simulations and experiments. The results of experiments showed that the timing discriminators were able to effectively improve the precision even if the detected distance changes.

From the Semantic Point Cloud to Heritage-Building Information Modeling: A Semiautomatic Approach Exploiting Machine Learning

Abstract: This work presents a semi-automatic approach to the 3D reconstruction of Heritage-Building Information Models from point clouds based on machine learning techniques The use of digital information systems leveraging on three-dimensional (3D) representations in architectural heritage documentation and analysis is ever increasing For the creation of such repositories, reality-based surveying techniques, such as photogrammetry and laser scanning, allow the fast collection of reliable digital replicas of the study objects in the form of point clouds Besides, their output is raw and unstructured, and the transition to intelligible and semantic 3D representations is still a scarcely automated and time-consuming process requiring considerable human intervention More refined methods for 3D data interpretation of heritage point clouds are therefore sought after In tackling these issues, the proposed approach relies on (i) the application of machine learning techniques to semantically label 3D heritage data by identification of relevant geometric, radiometric and intensity features, and (ii) the use of the annotated data to streamline the construction of Heritage-Building Information Modeling (H-BIM) systems, where purely geometric information derived from surveying is associated with semantic descriptors on heritage documentation and management The “Grand-Ducal Cloister” dataset, related to the emblematic case study of the Pisa Charterhouse, is discussed

Towards Automated 3D Inspection of Water Leakages in Shield Tunnel Linings Using Mobile Laser Scanning Data.

Abstract: On-site manual inspection of metro tunnel leakages has been faced with the problems of low efficiency and poor accuracy. An automated, high-precision, and robust water leakage inspection method is vital to improve the manual approach. Existing approaches cannot provide the leakage location due to the lack of spatial information. Therefore, an integrated deep learning method of water leakage inspection using tunnel lining point cloud data from mobile laser scanning is presented in this paper. It is composed of three parts as follows: (1) establishment of the water leakage dataset using the acquired point clouds of tunnel linings; (2) automated leakage detection via a mask-region-based convolutional neural network; and (3) visualization and quantitative evaluation of the water leakage in 3D space via a novel triangle mesh method. The testing result reveals that the proposed method achieves automated detection and evaluation of tunnel lining water leakages in 3D space, which provides the inspectors with an intuitive overall 3D view of the detected water leakages and the leakage information (area, location, lining segments, etc.).