• ETRI Journal
• /
• v.45 no.5
• /
• pp.847-861
• /
• 2023

Dynamic object detection is essential for ensuring safe and reliable autonomous driving. Recently, light detection and ranging (LiDAR)-based object detection has been introduced and shown excellent performance on various benchmarks. Although LiDAR sensors have excellent accuracy in estimating distance, they lack texture or color information and have a lower resolution than conventional cameras. In addition, performance degradation occurs when a LiDAR-based object detection model is applied to different driving environments or when sensors from different LiDAR manufacturers are utilized owing to the domain gap phenomenon. To address these issues, a sensor-fusion-based object detection and classification method is proposed. The proposed method operates in real time, making it suitable for integration into autonomous vehicles. It performs well on our custom dataset and on publicly available datasets, demonstrating its effectiveness in real-world road environments. In addition, we will make available a novel three-dimensional moving object detection dataset called ETRI 3D MOD.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.27 no.2
• /
• pp.189-196
• /
• 2024

Classification of drones and birds is challenging due to diverse flight patterns and limited data availability. Previous research has focused on identifying the flight patterns of unmanned aerial vehicles by emphasizing dynamic features such as speed and heading. However, this approach tends to neglect crucial spatial information, making accurate discrimination of unmanned aerial vehicle characteristics challenging. Furthermore, training methods for situations with imbalanced data among classes have not been proposed by traditional machine learning techniques. In this paper, we propose a data processing method that preserves angle information while maintaining positional details, enabling the deep learning model to better comprehend positional information of drones. Additionally, we introduce a training technique to address the issue of data imbalance.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.8 no.1
• /
• pp.61-68
• /
• 1988

This study is intended to propose an approach to reliability-based safety evaluation as well as LRFR(Load and Resisitance Factor Rating) type capacity classification of military or civilian bridges based on the limit state models which are delived by incorporating all the uncertainties of resistance and load random variables including deterioration, and are used in a practical AFOSM (Advanced First Order Second Moment) method. The proposed methods for the assement of safety and load carrying capacity are applied for the evaluation of rating and classifications of several practical bridges against the crossing of military vehicles. Based on the observation of the numerical results, it can be concluded that the current NATO classification method which is based on the traditionl allowable stress concept can not provide real load carrying capacity but results in nominal classification, and therefore the reliability-based safety evaluation and LRFR-classification method or the corresponding rational allowable stress method proposed in this paper may have to be introduced into the classification practice.

• Resources Recycling
• /
• v.27 no.2
• /
• pp.11-23
• /
• 2018

The need for remanufacturing automotive parts is required due to the depletion of resources, rising raw material prices and strengthening environmental regulations. For remanufacturing, stable supply and demand of core must be accompanied. At present, remanufacturing companies collect cores through various routes, but the recovery rate of cores from the End-of-Life Vehicles is low. If we can systematically collect cores from hundreds of thousands of ELVs that were generated each year, the recovery rate of the core for remanufacturing will be further improved. Therefore, in this paper, we tried to establish a classification system for the ELV as a method for collecting the cores from the ELV. First, we selected the elements affecting the classification and determined the scope for the evaluation. The final rating classification is established by calculating the weights among the influence elements. Finally, through the case study, the dismantling grade of the actual ELV was evaluated to derive the second grade.

• Journal of Advanced Navigation Technology
• /
• v.20 no.6
• /
• pp.604-611
• /
• 2016

Drone, which refers to unmanned aerial vehicles (UAV), industries are improving rapidly and exceeding existing level of remote controlled aircraft models. Also, they are applying automation and cloud network technology. Recently, the ability of drones can bring serious threats to public safety such as explosives and unmanned aircraft carrying hazardous materials. On the purpose of reducing these kinds of threats, it is necessary to detect these illegal drones, using acoustic feature extraction and classifying technology. In this paper, we introduce sound feature vector extraction method by harmonic feature extraction method (HLA). Feature vector extraction method based on HLA make it possible to distinguish drone sound, extracting features of sound data. In order to assess the performance of distinguishing sounds which exists in outdoor environment, we analyzed various sounds of things and real drones, and classified sounds of drone and others as simulation of each sound source.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.41 no.12
• /
• pp.1-7
• /
• 2004

We have developed a 24GHz side-looking vehicle detection radar. A 24GHz front-end module and a novel signal processing algorithm have been developed for speed measurement and size classification of vehicles in multiple lanes. The system has a fixed antenna and FMCW processing module. This paper presents the background theory of operation and shows some measured data using the algorithm. The data shows that measured velocity of the passing vehicle is within the accuracy of 95% in single lane and the velocity of the vehicles in two lanes is within the accuracy of 90% by using variable threshold estimation. The classification of vehicle size as small, medium and large has been measured with 89% accuracy.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.3
• /
• pp.160-174
• /
• 2023

LiDAR plays a key role in autonomous vehicles, and to improve its visibility, it is necessary to improve its performance and the detection objects. Accordingly, this study proposes a shape for traffic safety signs that is advantageous for self-driving vehicles to recognize. Improvement plans are also proposed using a shape-recognition algorithm based on point cloud data collected through LiDAR sensors. For the experiment, a DBSCAN-based road-sign recognition and classification algorithm, which is commonly used in point cloud research, was developed, and a 32ch LiDAR was used in an actual road environment to conduct recognition performance tests for 5 types of road signs. As a result of the study, it was possible to detect a smaller number of point clouds with a regular triangle or rectangular shape that has vertical asymmetry than a square or circle. The results showed a high classification accuracy of 83% or more. In addition, when the size of the square mark was enlarged by 1.5 times, it was possible to classify it as a square despite an increase in the measurement distance. These results are expected to be used to improve dedicated roads and traffic safety facilities for sensors in the future autonomous driving era and to develop new facilities.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.6
• /
• pp.284-298
• /
• 2022

It is necessary to acquire information on traffic safety facilities installed on the roadways specifically for the operation of autonomous vehicles. The purpose of this study is to prepare a Korean version of an integrated message-set design as a way to provide to autonomous vehicles standardized information on traffic safety facilities. In this study, necessary facilities are classified according to four criteria (no legal basis; not providing information to autonomous vehicles; providing duplicate information; not standardized, and too difficult to generalize) based on information that must be provided to operate autonomous vehicles. The priority of information delivery (gross negligence followed by behavior change) was classified according to the importance of the information to be provided during autonomous driving, and the form was defined for the classification code in the information delivered. Finally, the information location and delivery method of traffic facilities for compliance with SAE J2735 were identified. This study is meaningful in that it provides a plan for roadway operations by suggesting a method for providing information to autonomously driven vehicles.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.2
• /
• pp.217-223
• /
• 2014

Recently, various researches and projects about electric vehicles are in progress vigorously and continuously and it is expected to penetrate rapidly with the next a few years. This deployment will cause the change of load composition rate affecting on power system planning and operations. Therefore, a new load model should be developed integrating with electric vehicle loads. In this paper, the load composition rate of residential sectors is analyzed considering the deployment of this mobility load such as electric vehicles and a new diffusion model is proposed based on the classification of the replacement patterns. Additionally, electric vehicle charging loads are basically modeled by some individual load experiments to develop new load models for smart place and some new conceptual power systems such as micro grids.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.11
• /
• pp.1008-1016
• /
• 2015

This paper proposes a method for detecting the front side of vehicles. The method can find the car side with a license plate even with complicated and cluttered backgrounds. A convolutional neural network (CNN) is used to solve the detection problem as a unified framework combining feature detection, classification, searching, and localization estimation and improve the reliability of the system with simplicity of usage. The proposed CNN structure avoids sliding window search to find the locations of vehicles and reduces the computing time to achieve real-time processing. Multiple responses of the network for vehicle position are further processed by a weighted clustering and probabilistic threshold decision method. Experiments using real images in parking lots show the reliability of the method.