scholarly journals CNN based lane detection with instance segmentation in edge-cloud computing

2020 ◽  
Vol 9 (1) ◽  
Author(s):  
Wei Wang ◽  
Hui Lin ◽  
Junshu Wang
Keyword(s):  
Cloud Computing ◽  
Data Processing ◽  
Autonomous Driving ◽  
Lane Detection ◽  
Dual Model ◽  
High Definition ◽  
Processing Efficiency ◽  
Cloud Data ◽  
Lane Recognition ◽  
Instance Segmentation

Abstract At present, the number of vehicle owners is increasing, and the cars with autonomous driving functions have attracted more and more attention. The lane detection combined with cloud computing can effectively solve the drawbacks of traditional lane detection relying on feature extraction and high definition, but it also faces the problem of excessive calculation. At the same time, cloud data processing combined with edge computing can effectively reduce the computing load of the central nodes. The traditional lane detection method is improved, and the current popular convolutional neural network (CNN) is used to build a dual model based on instance segmentation. In the image acquisition and processing processes, the distributed computing architecture provided by edge-cloud computing is used to improve data processing efficiency. The lane fitting process generates a variable matrix to achieve effective detection in the scenario of slope change, which improves the real-time performance of lane detection. The method proposed in this paper has achieved good recognition results for lanes in different scenarios, and the lane recognition efficiency is much better than other lane recognition models.

Cloud Data Management Based on Cloud Computing Technology

2014 ◽  
Vol 543-547 ◽  
pp. 3573-3576
Author(s):  
Yuan Jun Zou
Keyword(s):  
Cloud Computing ◽  
Parallel Computing ◽  
Data Processing ◽  
Development Planning ◽  
Computer Data ◽  
Cloud Data ◽  
Computing Platform ◽  
Convergence Error ◽  
Cloud Computing Platform ◽  
Cloud Data Processing

Cloud computing, networking and other high-end computer data processing technology are the important contents of eleven-five development planning in China. They have developed rapidly in recent years in the field of engineering. In this paper, we combine parallel computing with the collaborative simulation principle, design a cloud computing platform, establish the mathematical model of cloud data processing and parallel computing algorithm, and verify the applicability of algorithm through the numerical simulation. Through numerical calculation, cloud computing platform can be divided into complex grids, and the transmission speed is fast, which is eight times than the finite difference method. The mesh is meticulous, which reaches millions. Convergence error is minimum, only 0.001. The calculation accuracy is up to 98.36%.

Security Issues in Fog Computing and ML-Based Solutions

2021 ◽  
pp. 209-234
Author(s):  
Himanshu Sahu ◽  
Gaytri
Keyword(s):  
Machine Learning ◽  
Cloud Computing ◽  
Data Processing ◽  
Fog Computing ◽  
Machine Learning Techniques ◽  
Special Focus ◽  
Cloud Data Center ◽  
Cloud Data ◽  
Security Issues ◽  
Learning Techniques

IoT requires data processing, which is provided by the cloud and fog computing. Fog computing shifts centralized data processing from the cloud data center to the edge, thereby supporting faster response due to reduced communication latencies. Its distributed architecture raises security and privacy issues; some are inherited from the cloud, IoT, and network whereas others are unique. Securing fog computing is equally important as securing cloud computing and IoT infrastructure. Security solutions used for cloud computing and IoT are similar but are not directly applicable in fog scenarios. Machine learning techniques are useful in security such as anomaly detection, intrusion detection, etc. So, to provide a systematic study, the chapter will cover fog computing architecture, parallel technologies, security requirements attacks, and security solutions with a special focus on machine learning techniques.

Point Cloud Data Preprocessing Based on the Geomagic

2013 ◽  
Vol 690-693 ◽  
pp. 2817-2820 ◽  
Author(s):  
Feng Chang Xue
Keyword(s):  
Data Processing ◽  
Noise Reduction ◽  
Point Cloud ◽  
Point Cloud Data ◽  
Processing Efficiency ◽  
Cloud Data ◽  
Noise Point ◽  
Isolated Point ◽  
Cloud Process

In the acquisition process of point cloud data, there exist speckles and noise point. This essay, based on geomagic, presents specific technical steps to remove isolated point in vitro, avoid point cloud resampling and remove point cloud noise. Applications showed that geomagic can, from the original point cloud process, find out useful point cloud, delete speckles, and through the methods of resampling and noise reduction processing, improve cloud quality and reduce data deviation thus improving the data processing efficiency.

scholarly journals Assessment of Benefits and Disadvantages of Implementing Cloud-Specific Solutions in Polish Companies on the Example of ERP Systems

2020 ◽  
Author(s):  
Remigiusz Zieliński ◽  
Sebastian Kot ◽  
Katarzyna Zielińska
Keyword(s):  
Cloud Computing ◽  
Data Processing ◽  
Operational Efficiency ◽  
Erp Systems ◽  
Cloud Data ◽  
It Systems ◽  
Complex Process ◽  
Questionnaire Surveys ◽  
Cloud Data Processing ◽  
Theoretical Considerations

Abstract Both dynamics and environmental turbulences result in constant updating of IT and communication specific technologies utilized within the scope of various organizations. The Polish IT market is characterized by a dynamic development of the cloud data processing model that offers access to various resources via the network in the form of convenient services. The transfer of IT systems (including ERP) to the cloud is a complex process and while the connection can be made, it may be met with a number of problems. The predominant aim of the article, aside from theoretical considerations, is to showcase the possibilities of supporting a company’s operational schemes by means of IT solutions available within the cloud; as well as to indicate both the benefits and the difficulties Polish companies have to face while implementing cloud computing specific solutions. Within the scope of the article, selected outcomes of questionnaire surveys are going to be presented. Research results have shown that IT solutions and ERP systems available in the cloud have been positively perceived by the respondents, especially with regard to improving the overall operational efficiency of companies.

Optimization and Management in Order to Drive Targeted Networking Memory Database

2015 ◽  
Vol 727-728 ◽  
pp. 965-968
Author(s):  
Ji Liu
Keyword(s):  
Cloud Computing ◽  
Wireless Communication ◽  
Data Processing ◽  
Real Time ◽  
System Architecture ◽  
Large Data ◽  
Processing Efficiency ◽  
Computing Platform ◽  
Cloud Computing Platform ◽  
Vehicle Networking

In today's vehicle networking system architecture is mainly composed of four parts: sensor networks, wireless communication networks, cloud computing platforms and vehicle terminal. Wireless sensor network is responsible for the front of the real-time collection of traffic information, a wireless communication network to send information to the backend of the cloud computing platform, cloud computing platform to handle a large number of vehicles to collect real-time information from the front, and finally sends the information to the end user. In this thesis, this car networking research background, analyze vehicle networking system architecture consisting of performance indicators for each part of the system recognize cloud platform for large data processing efficiency as well as room for improvement. Then put forward the traditional computing platform I / O disk database with in-memory database to replace the cloud to enhance cloud computing platform for large data processing efficiency.

scholarly journals RS-Lane: A Robust Lane Detection Method Based on ResNeSt and Self-Attention Distillation for Challenging Traffic Situations

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Ronghui Zhang ◽  
Yueying Wu ◽  
Wanting Gou ◽  
Junzhou Chen
Keyword(s):  
Detection Method ◽  
Autonomous Driving ◽  
Lane Detection ◽  
Feature Representation ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Advanced Driver Assistance Systems ◽  
Assistance Systems ◽  
Light And Shadow ◽  
Instance Segmentation

Lane detection plays an essential part in advanced driver-assistance systems and autonomous driving systems. However, lane detection is affected by many factors such as some challenging traffic situations. Multilane detection is also very important. To solve these problems, we proposed a lane detection method based on instance segmentation, named RS-Lane. This method is based on LaneNet and uses Split Attention proposed by ResNeSt to improve the feature representation on slender and sparse annotations like lane markings. We also use Self-Attention Distillation to enhance the feature representation capabilities of the network without adding inference time. RS-Lane can detect lanes without number limits. The tests on TuSimple and CULane datasets show that RS-Lane has achieved comparable results with SOTA and has improved in challenging traffic situations such as no line, dazzle light, and shadow. This research provides a reference for the application of lane detection in autonomous driving and advanced driver-assistance systems.

scholarly journals Automated Extraction of Driving Lines from Mobile Laser Scanning Point Clouds

2019 ◽  
Vol 1 ◽  
pp. 1-6
Author(s):  
Lingfei Ma ◽  
Tianyu Wu ◽  
Ying Li ◽  
Jonathan Li ◽  
Yiping Chen ◽  
...  
Keyword(s):  
Curve Fitting ◽  
Laser Scanning ◽  
Region Growing ◽  
Point Clouds ◽  
Autonomous Driving ◽  
Road Surface ◽  
High Definition ◽  
Localization Accuracy ◽  
Cloud Data ◽  
Road Marking

<p><strong>Abstract.</strong> This paper presents a novel approach to automated generation of driving lines from mobile laser scanning (MLS) point cloud data. The proposed method consists of three steps: road surface segmentation, road marking extraction and classification, and driving line generation. The voxel-based upward-growing algorithm was firstly used to extract ground points from the raw MLS point clouds followed by segmentation of road surface using a region-growing algorithm. Then, the statistical outlier removal filter was applied to separate and refine the road marking points followed by extracting and classifying the lane markings based on the geometric features of different road markings using empirical hierarchical decision trees. Finally, land node structures were constructed followed by generation of driving lines using a curve-fitting algorithm. The proposed method was tested on both circular road sections and irregular intersections. The smoothing spline curve fitting model was tested on the circular road sections, while the Catmull-Rom spline with five control points was used to generate the driving lines at road intersections. The overall performance of the proposed algorithms is promising with 96.0% recall, 100.0% precision, and 98.0% F1-score for the lane marking extraction specifically. Most significantly, the validation results demonstrate that the driving lines can be effectively generated within 20&amp;thinsp;cm-level localization accuracy at an average of 3.5% miscoding using MLS point clouds, which meets the requirement of localization accuracy of fully autonomous driving functions. The results demonstrate the proposed methods can successfully generate road driving lines in the test datasets to support the development of high-definition maps.</p>

A Group Collaboratable Proof of Retrievability Scheme for Cloud Data Storage

2014 ◽  
Vol 13 (7) ◽  
pp. 4625-4632
Author(s):  
Jyh-Shyan Lin ◽  
Kuo-Hsiung Liao ◽  
Chao-Hsing Hsu
Keyword(s):  
Cloud Computing ◽  
Data Storage ◽  
Cloud Service ◽  
Third Party ◽  
Group Collaboration ◽  
Cloud Data ◽  
Cloud Data Storage ◽  
Important Trend ◽  
Data Update ◽  
Proof Of Retrievability

Cloud computing and cloud data storage have become important applications on the Internet. An important trend in cloud computing and cloud data storage is group collaboration since it is a great inducement for an entity to use a cloud service, especially for an international enterprise. In this paper we propose a cloud data storage scheme with some protocols to support group collaboration. A group of users can operate on a set of data collaboratively with dynamic data update supported. Every member of the group can access, update and verify the data independently. The verification can also be authorized to a third-party auditor for convenience.

scholarly journals A Set of Single YOLO Modalities to Detect Occluded Entities via Viewpoint Conversion

2021 ◽  
Vol 11 (13) ◽  
pp. 6016
Author(s):  
Jinsoo Kim ◽  
Jeongho Cho
Keyword(s):  
Object Detection ◽  
Autonomous Vehicles ◽  
Autonomous Driving ◽  
Detection Algorithm ◽  
Detection Accuracy ◽  
Cloud Data ◽  
Detection Techniques ◽  
Bounding Boxes ◽  
Partially Occluded ◽  
Rgb Image

For autonomous vehicles, it is critical to be aware of the driving environment to avoid collisions and drive safely. The recent evolution of convolutional neural networks has contributed significantly to accelerating the development of object detection techniques that enable autonomous vehicles to handle rapid changes in various driving environments. However, collisions in an autonomous driving environment can still occur due to undetected obstacles and various perception problems, particularly occlusion. Thus, we propose a robust object detection algorithm for environments in which objects are truncated or occluded by employing RGB image and light detection and ranging (LiDAR) bird’s eye view (BEV) representations. This structure combines independent detection results obtained in parallel through “you only look once” networks using an RGB image and a height map converted from the BEV representations of LiDAR’s point cloud data (PCD). The region proposal of an object is determined via non-maximum suppression, which suppresses the bounding boxes of adjacent regions. A performance evaluation of the proposed scheme was performed using the KITTI vision benchmark suite dataset. The results demonstrate the detection accuracy in the case of integration of PCD BEV representations is superior to when only an RGB camera is used. In addition, robustness is improved by significantly enhancing detection accuracy even when the target objects are partially occluded when viewed from the front, which demonstrates that the proposed algorithm outperforms the conventional RGB-based model.

Sign in / Sign up

Export Citation Format

Share Document