A Tensor voting Based Surface Anomaly Classification Approach by Using 3D Point Cloud Data

2021 ◽  
pp. 1-41
Author(s):  
Juan Du ◽  
Hao Yan ◽  
Tzyy-Shuh Chang ◽  
Jianjun Shi
Keyword(s):  
Point Cloud ◽  
Data Representation ◽  
Complex Data ◽  
Tensor Voting ◽  
3D Point Cloud ◽  
Point Cloud Data ◽  
Cloud Data ◽  
Dimension Measurement ◽  
Anomaly Classification ◽  
Part Dimension

Abstract Advanced 3D scanning technology has been widely used in many industries to collect the massive point cloud data of artifacts for part dimension measurement and shape analysis. Though point cloud data also has product surface quality information, it is challenging to conduct effective surface anomaly classification due to the complex data representation, high-dimensionality, and inconsistent size of the 3D point cloud data within each sample. To deal with these challenges, this paper proposes a tensor voting based approach for anomaly classification of artifact surfaces. A case study based on 3D scanned data obtained from a manufacturing plant shows the effectiveness of the proposed method.

Fast Intent Prediction of Multi-Cyclists in 3D Point Cloud Data using Deep Neural Networks

2021 ◽  
Author(s):  
Khaled Saleh ◽  
Ahmed Abobakr ◽  
Mohammed Hossny ◽  
Darius Nahavandi ◽  
Julie Iskander ◽  
...  
Keyword(s):  
Neural Networks ◽  
Point Cloud ◽  
Deep Neural Networks ◽  
3D Point Cloud ◽  
Point Cloud Data ◽  
Cloud Data ◽  
Intent Prediction

Automated Surface Inspection Using 3D Point Cloud Data in Manufacturing: A Case Study

2018 ◽  
Author(s):  
Romina Dastoorian ◽  
Ahmad E. Elhabashy ◽  
Wenmeng Tian ◽  
Lee J. Wells ◽  
Jaime A. Camelio
Keyword(s):  
Point Cloud ◽  
Surface Defect ◽  
Fault Location ◽  
Three Dimensional ◽  
Practical Case ◽  
3D Point Cloud ◽  
Point Cloud Data ◽  
Defect Inspection ◽  
Cloud Data

With the latest advancements in three-dimensional (3D) measurement technologies, obtaining 3D point cloud data for inspection purposes in manufacturing is becoming more common. While 3D point cloud data allows for better inspection capabilities, their analysis is typically challenging. Especially with unstructured 3D point cloud data, containing coordinates at random locations, the challenges increase with higher levels of noise and larger volumes of data. Hence, the objective of this paper is to extend the previously developed Adaptive Generalized Likelihood Ratio (AGLR) approach to handle unstructured 3D point cloud data used for automated surface defect inspection in manufacturing. More specifically, the AGLR approach was implemented in a practical case study to inspect twenty-seven samples, each with a unique fault. These faults were designed to cover an array of possible faults having three different sizes, three different magnitudes, and located in three different locations. The results show that the AGLR approach can indeed differentiate between non-faulty and a varying range of faulty surfaces while being able to pinpoint the fault location. This work also serves as a validation for the previously developed AGLR approach in a practical scenario.

Assessing the Geometric Integrity of Additive Manufactured Parts From Point Cloud Data Using Spectral Graph Theoretic Sparse Representation-Based Classification

2017 ◽  
Author(s):  
M. Samie Tootooni ◽  
Ashley Dsouza ◽  
Ryan Donovan ◽  
Prahalad K. Rao ◽  
Zhenyu (James) Kong ◽  
...  
Keyword(s):  
Sparse Representation ◽  
Point Cloud ◽  
Extrusion Temperature ◽  
3D Point Cloud ◽  
Point Cloud Data ◽  
Modeling Framework ◽  
Cloud Data ◽  
Integrity Assessment ◽  
Graph Theoretic ◽  
Spectral Graph

This work proposes a novel approach for geometric integrity assessment of additive manufactured (AM, 3D printed) components, exemplified by acrylonitrile butadiene styrene (ABS) polymer parts made using fused filament fabrication (FFF) process. The following two research questions are addressed in this paper: (1) what is the effect of FFF process parameters, specifically, infill percentage (If) and extrusion temperature (Te) on geometric integrity of ABS parts?; and (2) what approach is required to differentiate AM parts with respect to their geometric integrity based on sparse sampling from a large (∼ 2 million data points) laser-scanned point cloud dataset? To answer the first question, ABS parts are produced by varying two FFF parameters, namely, infill percentage (If) and extrusion temperature (Te) through design of experiments. The part geometric integrity is assessed with respect to key geometric dimensioning and tolerancing (GD&T) features, such as flatness, circularity, cylindricity, root mean square deviation, and in-tolerance percentage. These GD&T parameters are obtained by laser scanning of the FFF parts. Concurrently, coordinate measurements of the part geometry in the form of 3D point cloud data is also acquired. Through response surface statistical analysis of this experimental data it was found that discrimination of geometric integrity between FFF parts based on GD&T parameters and process inputs alone was unsatisfactory (regression R2 < 50%). This directly motivates the second question. Accordingly, a data-driven analytical approach is proposed to classify the geometric integrity of FFF parts using minimal number (< 2% of total) of laser-scanned 3D point cloud data. The approach uses spectral graph theoretic Laplacian eigenvalues extracted from the 3D point cloud data in conjunction with a modeling framework called sparse representation to classify FFF part quality contingent on the geometric integrity. The practical outcome of this work is a method that can quickly classify the part geometric integrity with minimal point cloud data and high classification fidelity (F-score > 95%), which bypasses tedious coordinate measurement.

scholarly journals Modeling and Processing of Smart Point Clouds of Cultural Relics with Complex Geometries

2021 ◽  
Vol 10 (9) ◽  
pp. 617
Author(s):  
Su Yang ◽  
Miaole Hou ◽  
Ahmed Shaker ◽  
Songnian Li
Keyword(s):  
Point Cloud ◽  
Semantic Information ◽  
Point Clouds ◽  
Information Modeling ◽  
Complex Geometries ◽  
3D Point Cloud ◽  
Point Cloud Data ◽  
Modeling Framework ◽  
Cloud Data ◽  
Cultural Relics

The digital documentation of cultural relics plays an important role in archiving, protection, and management. In the field of cultural heritage, three-dimensional (3D) point cloud data is effective at expressing complex geometric structures and geometric details on the surface of cultural relics, but lacks semantic information. To elaborate the geometric information of cultural relics and add meaningful semantic information, we propose a modeling and processing method of smart point clouds of cultural relics with complex geometries. An information modeling framework for complex geometric cultural relics was designed based on the concept of smart point clouds, in which 3D point cloud data are organized through the time dimension and different spatial scales indicating different geometric details. The proposed model allows smart point clouds or a subset to be linked with semantic information or related documents. As such, this novel information modeling framework can be used to describe rich semantic information and high-level details of geometry. The proposed information model not only expresses the complex geometric structure of the cultural relics and the geometric details on the surface, but also has rich semantic information, and can even be associated with documents. A case study of the Dazu Thousand-Hand Bodhisattva Statue, which is characterized by a variety of complex geometries, reveals that our proposed framework is capable of modeling and processing the statue with excellent applicability and expansibility. This work provides insights into the sustainable development of cultural heritage protection globally.

Sign in / Sign up

Export Citation Format

Share Document