scholarly journals The Fast and Continuous Railway Clearance Inspection Algorithm Based on Vehicle LiDAR Point Cloud

2020 ◽  
Author(s):  
Dan Zhong ◽  
Tonggang Zhang ◽  
Yuhui Kan ◽  
Fugui Xie
Keyword(s):  
Point Cloud ◽  
Experimental Results ◽  
Problem Of Time ◽  
Axial Alignment ◽  
Bounding Box ◽  
Accurate Judgment ◽  
Safety Operation ◽  
Train Safety

Railway clearance inspection is the foundation work to ensure train safety operation. Aiming at the problem of time-consuming and low accuracy of clearance inspection based on vehicle LiDAR point cloud, this paper proposes a fast and continuous railway clearance inspection algorithm. Firstly, the preliminary clearance inspection is completed by constructing the axial alignment bounding box of the clearance polyhedron. Then fine clearance inspection based on the topology relationship between point cloud and polyhedron to provide an accurate judgment. Finally, the clearance distribution information is obtained by Euclidean clustering. Based on three representative datasets, the proposed algorithm is validated, and the experimental results show that the accuracy and efficiency are enhanced by 25.0% and 83.4% responsibly.

A Configurable Tilt License Plate Correction Method Based on Parallel Lines

2013 ◽  
Vol 385-386 ◽  
pp. 1429-1433 ◽  
Author(s):  
Zhong Yan Liang ◽  
San Yuan Zhang
Keyword(s):  
Recognition System ◽  
Correction Method ◽  
Selection Method ◽  
Experimental Results ◽  
License Plate ◽  
License Plate Recognition ◽  
Parallel Lines ◽  
Bounding Box ◽  
Bounding Boxes ◽  
Straight Lines

The tilt license plate correction is an important part of the license plate recognition system. Traditional correction methods are based on one theory. It is difficult to use the advantages of different approaches. We propose some methods to help improve the tile license plate correction: a bounding box selection method based on similar height and a mutual correction method based on fitted parallel straight lines. Moreover, we use wide bounding boxes to segment touched characters. If the method based on parallel lines fails, another method, such as PCA-based one, can be used for complement. Experimental results show the proposed method outperforms others.

scholarly journals AUTOMATIC CLUSTERING OF CELTIC COINS BASED ON 3D POINT CLOUD PATTERN ANALYSIS

2020 ◽  
Vol V-2-2020 ◽  
pp. 973-980
Author(s):  
S. Horache ◽  
F. Goulette ◽  
J.-E. Deschaud ◽  
T. Lejars ◽  
K. Gruel
Keyword(s):  
Point Cloud ◽  
Pattern Analysis ◽  
Experimental Results ◽  
3D Point Cloud ◽  
Automatic Clustering ◽  
Clustering Quality ◽  
3D Scans ◽  
Cloud Pattern ◽  
Graph Based Clustering

Abstract. The recognition and clustering of coins which have been struck by the same die is of interest for archeological studies. Nowadays, this work can only be performed by experts and is very tedious. In this paper, we propose a method to automatically cluster dies, based on 3D scans of coins. It is based on three steps: registration, comparison and graph-based clustering. Experimental results on 90 coins coming from a Celtic treasury from the II-Ith century BC show a clustering quality equivalent to expert’s work.

scholarly journals Experiments on learning-based industrial bin-picking with iterative visual recognition

2018 ◽  
Vol 45 (4) ◽  
pp. 446-457 ◽  
Author(s):  
Kensuke Harada ◽  
Weiwei Wan ◽  
Tokuo Tsuji ◽  
Kohei Kikuchi ◽  
Kazuyuki Nagata ◽  
...  
Keyword(s):  
Random Forest ◽  
Point Cloud ◽  
Visual Recognition ◽  
Experimental Results ◽  
Training Data ◽  
Robotic Assembly ◽  
Depth Sensor ◽  
Content Type ◽  
New Finding ◽  
Bin Picking

Purpose This paper aims to automate the picking task needed in robotic assembly. Parts supplied to an assembly process are usually randomly staked in a box. If randomized bin-picking is introduced to a production process, we do not need any part-feeding machines or human workers to once arrange the objects to be picked by a robot. The authors introduce a learning-based method for randomized bin-picking. Design/methodology/approach The authors combine the learning-based approach on randomized bin-picking (Harada et al., 2014b) with iterative visual recognition (Harada et al., 2016a) and show additional experimental results. For learning, we use random forest explicitly considering the contact between a finger and a neighboring object. The iterative visual recognition method iteratively captures point cloud to obtain more complete point cloud of piled object by using 3D depth sensor attached at the wrist. Findings Compared with the authors’ previous research (Harada et al., 2014b) (Harada et al., 2016a), their new finding is as follows: by using random forest, the number of training data becomes extremely small. By adding penalty to occluded area, the learning-based method predicts the success after point cloud with less occluded area. We analyze the calculation time of the iterative visual recognition. We furthermore make clear the cases where a finger contacts neighboring objects. Originality/value The originality exists in the part where the authors combined the learning-based approach with the iterative visual recognition and supplied additional experimental results. After obtaining the complete point cloud of the piled object, prediction becomes effective.

scholarly journals Weakly Supervised Learning for Object Localization Based on an Attention Mechanism

2021 ◽  
Vol 11 (22) ◽  
pp. 10953
Author(s):  
Nojin Park ◽  
Hanseok Ko
Keyword(s):  
Deep Learning ◽  
Experimental Results ◽  
Object Localization ◽  
Localization Accuracy ◽  
Large Size ◽  
Bounding Box ◽  
Improved Performance ◽  
Weakly Supervised ◽  
Detection And Localization ◽  
Activation Map

Recently, deep learning has been successfully applied to object detection and localization tasks in images. When setting up deep learning frameworks for supervised training with large datasets, strongly labeling the objects facilitates good performance; however, the complexity of the image scene and large size of the dataset make this a laborious task. Hence, it is of paramount importance that the expensive work associated with the tasks involving strong labeling, such as bounding box annotation, is reduced. In this paper, we propose a method to perform object localization tasks without bounding box annotation in the training process by means of employing a two-path activation-map-based classifier framework. In particular, we develop an activation-map-based framework to judicially control the attention map in the perception branch by adding a two-feature extractor so that better attention weights can be distributed to induce improved performance. The experimental results indicate that our method surpasses the performance of the existing deep learning models based on weakly supervised object localization. The experimental results show that the proposed method achieves the best performance, with 75.21% Top-1 classification accuracy and 55.15% Top-1 localization accuracy on the CUB-200-2011 dataset.

scholarly journals Acquisition and Registration of Human Point Cloud Based on Depth Camera

2018 ◽  
Vol 173 ◽  
pp. 03001
Author(s):  
Huibai Wang ◽  
Shengyu Wang
Keyword(s):  
Point Cloud ◽  
Feature Matching ◽  
Least Square ◽  
Experimental Results ◽  
Depth Camera ◽  
Point Cloud Registration ◽  
Camera Pose ◽  
Cloud Points

Registration of human cloud points is a key step in 3D human reconstruction. So, this paper proposes using depth camera to acquire human point cloud. Then, using feature matching and PnP (Perspective-n-Point) to solve the camera pose.Finally, use least square to optimize camera pose and implement point cloud registration. Experimental results show the effectiveness of this method.

Research on the integrated manipulator of point cloud measurement and precise cutting for waste nuclear tank

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Ruolong Qi ◽  
Wenfeng Liang
Keyword(s):  
Data Structure ◽  
Nuclear Waste ◽  
Point Cloud ◽  
Cloud Model ◽  
Processing System ◽  
Machining Accuracy ◽  
Content Type ◽  
Bounding Box ◽  
Calibration Methods ◽  
Waste Tanks

Purpose Nuclear waste tanks need to be cut into pieces before they can be safely disposed of, but the cutting process produces a large amount of aerosols with radiation, which is very harmful to the health of the operator. The purpose of this paper is to establish an intelligent strategy for an integrated robot designed for measurement and cutting, which can accurately identify and cut unknown nuclear waste tanks and realize autonomous precise processing. Design/methodology/approach A robot system integrating point cloud measurement and plasma cutting is designed in this paper. First, accurate calibration methods for the robot, tool and hand-eye system are established. Second, for eliminating the extremely scattered point cloud caused by metal surface refraction, an omnidirectional octree data structure with 26 vectors is proposed to extract the point cloud model more accurately. Then, a minimum bounding box is calculated for limiting the local area to be cut, the local three-dimensional shape of the nuclear tank is fitted within the bounding box, in which the cutting trajectories and normal vectors are planned accurately. Findings The cutting precision is verified by changing the tool into a dial indicator in the simulation and the experiment process. The octree data structure with omnidirectional pointing vectors can effectively improve the filtering accuracy of the scattered point cloud. The point cloud filter algorithm combined with the structure calibration methods for the integrated measurement and processing system can ensure the final machining accuracy of the robot. Originality/value Aiming at the problems of large measurement noise interference, complex transformations between coordinate systems and difficult accuracy guarantee, this paper proposes structure calibration, point cloud filtering and point cloud-based planning algorithm, which can greatly improve the reliability and accuracy of the system. Simulation and experiment verify the final cutting accuracy of the whole system.

Point Cloud Reconstruction Based on SVR

2011 ◽  
Vol 403-408 ◽  
pp. 3267-3270
Author(s):  
Jin Guang Sun ◽  
Jun Tao Wang ◽  
Xin Nian Yang ◽  
Yang Li
Keyword(s):  
Support Vector Regression ◽  
Point Cloud ◽  
Reconstruction Algorithm ◽  
Surface Expression ◽  
Experimental Results ◽  
Support Vector ◽  
Implicit Function ◽  
Point Cloud Data ◽  
Cloud Data ◽  
Marching Cube Algorithm

This paper presents a point cloud reconstruction algorithm which based on SVR(support vector regression) . Firstly, the point cloud data pre-processing, filter out noise points. Then train the point by SVR , and we can get the function of surface expression. Finally, using the Marching Cube algorithm to visualize the implicit function. Experimental results show that the algorithm is more robust and more efficient.

scholarly journals Joint Geometry and Color Projection-based Point Cloud Quality Metric

2021 ◽  
Author(s):  
Alireza Javaheri ◽  
Catarina Brites ◽  
Fernando Pereira ◽  
Joao Ascenso
Keyword(s):  
Point Cloud ◽  
Pearson Correlation ◽  
Quality Metrics ◽  
Experimental Results ◽  
Reliable Estimate ◽  
Quality Metric ◽  
Joint Geometry ◽  
Objective Quality ◽  
Correlation Performance

Point cloud coding solutions have been recently standardized to address the needs of multiple application scenarios. The design and assessment of point cloud coding methods require reliable objective quality metrics to evaluate the level of degradation introduced by compression or any other type of processing. Several point cloud objective quality metrics has been recently proposed to reliable estimate human perceived quality, including the so-called projection-based metrics. In this context, this paper proposes a joint geometry and color projection-based point cloud objective quality metric which solves the critical weakness of this type of quality metrics, i.e., the misalignment between the reference and degraded projected images. Moreover, the proposed point cloud quality metric exploits the best performing 2D quality metrics in the literature to assess the quality of the projected images. The experimental results show that the proposed projection-based quality metric offers the best subjective-objective correlation performance in comparison with other metrics in the literature. The Pearson correlation gains regarding D1-PSNR and D2-PSNR metrics are 17% and 14.2 when data with all coding degradations is considered.

Sign in / Sign up

Export Citation Format

Share Document