Data Processing of Point Clouds for Object Detection for Structural Engineering Applications

Object detection in 3D point clouds is still a challenging task in autonomous driving. Due to the inherent occlusion and density changes of the point cloud, the data distribution of the same object will change dramatically. Especially, the incomplete data with sparsity or occlusion can not represent the complete characteristics of the object. In this paper, we proposed a novel strong–weak feature alignment algorithm between complete and incomplete objects for 3D object detection, which explores the correlations within the data. It is an end-to-end adaptive network that does not require additional data and can be easily applied to other object detection networks. Through a complete object feature extractor, we achieve a robust feature representation of the object. It serves as a guarding feature to help the incomplete object feature generator to generate effective features. The strong–weak feature alignment algorithm reduces the gap between different states of the same object and enhances the ability to represent the incomplete object. The proposed adaptation framework is validated on the KITTI object benchmark and gets about 6% improvement in detection average precision on 3D moderate difficulty compared to the basic model. The results show that our adaptation method improves the detection performance of incomplete 3D objects.

Download Full-text

Mechanical and X ray computed tomography characterisation of a WAAM 3D printed steel plate for structural engineering applications

Construction and Building Materials ◽

10.1016/j.conbuildmat.2020.121700 ◽

2020 ◽

pp. 121700

Author(s):

Z. Al-Nabulsi ◽

J.T. Mottram ◽

M. Gillie ◽

N. Kourra ◽

M.A. Williams

Keyword(s):

Computed Tomography ◽

Structural Engineering ◽

Steel Plate ◽

Engineering Applications ◽

X Ray ◽

3D Printed ◽

X Ray Computed

Download Full-text

Real-Time 3D object detection using improved convolutional neural network based on image-driven point cloud

Recent Advances in Electrical & Electronic Engineering (Formerly Recent Patents on Electrical & Electronic Engineering) ◽

10.2174/2352096514666211026142721 ◽

2021 ◽

Vol 14 ◽

Author(s):

Zhiyong Gao ◽

Jianhong Xiang

Keyword(s):

Neural Network ◽

Object Detection ◽

Convolutional Neural Network ◽

Real Time ◽

Point Cloud ◽

Point Clouds ◽

3D Point Cloud ◽

3D Object ◽

3D Object Detection ◽

Instance Segmentation

Background: While detecting the object directly from the 3D point cloud, the natural 3D patterns and invariance of 3D data are often obscure. Objective: In this work, we aimed at studying the 3D object detection from discrete, disordered and sparse 3D point clouds. Methods: The CNN is composed of the frustum sequence module, 3D instance segmentation module S-NET, 3D point cloud transformation module T-NET, and 3D boundary box estimation module E-NET. The search space of the object is determined by the frustum sequence module. The instance segmentation of the point cloud is performed by the 3D instance segmentation module. The 3D coordinates of the object are confirmed by the transformation module and the 3D bounding box estimation module. Results: Evaluated on KITTI benchmark dataset, our method outperforms the state of the art by remarkable margins while having real-time capability. Conclusion: We achieve real-time 3D object detection by proposing an improved convolutional neural network (CNN) based on image-driven point clouds.

Download Full-text

Adversarial Training on Point Clouds for Sim-to-Real 3D Object Detection

IEEE Robotics and Automation Letters ◽

10.1109/lra.2021.3093869 ◽

2021 ◽

pp. 1-1

Author(s):

Robert Debortoli ◽

Fuxin Li ◽

Ashish Kapoor ◽

Geoffrey Hollinger

Keyword(s):

Object Detection ◽

Point Clouds ◽

3D Object ◽

Adversarial Training ◽

3D Object Detection

Download Full-text

RECONSTRUCTION OF INDOOR MODELS USING POINT CLOUDS GENERATED FROM SINGLE-LENS REFLEX CAMERAS AND DEPTH IMAGES

ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences ◽

10.5194/isprsarchives-xl-4-w5-99-2015 ◽

2015 ◽

Vol XL-4/W5 ◽

pp. 99-102

Author(s):

F. Tsai ◽

T.-S. Wu ◽

I.-C. Lee ◽

H. Chang ◽

A. Y. S. Su

Keyword(s):

Data Processing ◽

Data Acquisition ◽

Three Dimensional ◽

Point Clouds ◽

Data Acquisition System ◽

Acquisition System ◽

Indoor Environments ◽

Single Lens ◽

Processing Procedure ◽

Data Processing Procedure

This paper presents a data acquisition system consisting of multiple RGB-D sensors and digital single-lens reflex (DSLR) cameras. A systematic data processing procedure for integrating these two kinds of devices to generate three-dimensional point clouds of indoor environments is also developed and described. In the developed system, DSLR cameras are used to bridge the Kinects and provide a more accurate ray intersection condition, which takes advantage of the higher resolution and image quality of the DSLR cameras. Structure from Motion (SFM) reconstruction is used to link and merge multiple Kinect point clouds and dense point clouds (from DSLR color images) to generate initial integrated point clouds. Then, bundle adjustment is used to resolve the exterior orientation (EO) of all images. Those exterior orientations are used as the initial values to combine these point clouds at each frame into the same coordinate system using Helmert (seven-parameter) transformation. Experimental results demonstrate that the design of the data acquisition system and the data processing procedure can generate dense and fully colored point clouds of indoor environments successfully even in featureless areas. The accuracy of the generated point clouds were evaluated by comparing the widths and heights of identified objects as well as coordinates of pre-set independent check points against in situ measurements. Based on the generated point clouds, complete and accurate three-dimensional models of indoor environments can be constructed effectively.

Download Full-text