• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.9
• /
• pp.1088-1095
• /
• 2019

The information and communication technology that is being developed recently has been greatly influencing the automobile market. In recent years, devices equipped with IT technology have been installed for the safety and convenience of the driver. However, it has the advantage of increased convenience as well as the disadvantage of increasing traffic accidents due to driver's distraction. In order to prevent such accidents, it is necessary to develop safety systems of various types and ways. In this paper implements a platform that can recognize LDWS and FCWS and PDWS by using a single camera without using radar sensor and camera fusion and stereo camera method using two or more sensors, and proposes to study multi-function driving safety platform using a single camera by analyzing recognition rate evaluation and validity on a vehicle.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.4
• /
• pp.699-708
• /
• 2021

In this paper, a steering control and speed control algorithm was presented for autonomous driving based on two vision sensors and road speed sign board. A car speed control algorithm was developed to recognize the speed sign by using TensorFlow, a deep learning program provided by Google to the road speed sign image provided from vision sensor B, and then let the car follows the recognized speed. At the same time, a steering angle control algorithm that detects lanes by analyzing road images transmitted from vision sensor A in real time, calculates steering angles, controls the front axle through PWM control, and allows the vehicle to track the lane. To verify the effectiveness of the proposed algorithm's steering and speed control algorithms, a car's prototype based on the Python language, Raspberry Pi and OpenCV was made. In addition, accuracy could be confirmed by verifying various scenarios related to steering and speed control on the test produced track.

• The Journal of the Korea Contents Association
• /
• v.21 no.5
• /
• pp.868-876
• /
• 2021

A school zone means a certain section designated among facilities, such as schools, childcare facilities, academies, etc., to protect children from the risk of traffic accidents. Since the school zone was established in September 1995, school zone accidents have not decreased even though it has been strengthened through a total of nine revisions until January 2021. This paper aims to present a standard model for child protection zones based on disaster prevention design. Methods of research included literature research, empirical research, and cognition research. Awareness survey was conducted on children, parents and drivers. Environmental surveys included crosswalks, motorways, pedestrian roads, and traffic lights. The investigation found that visual recognition of school zones by vehicle drivers was difficult, and found that motorcycles using pedestrian roads were a very threat to children. Accordingly, improved school zone standard model design centered on disaster prevention design was presented.

• Journal of Drive and Control
• /
• v.17 no.4
• /
• pp.23-30
• /
• 2020

Driving on irregular and inclined roads using agricultural machinery such as spraying machines or trucks in orchards causes farmer casualties associated with the overturning of agricultural machinery. In addition, the harm to agricultural workers caused by the excessive inhalation of the scattered pesticide frequently occurs during pest control processes. To address these problems, we introduced precision agricultural technology that could selectively spray pesticides only where the fruit is present by recognizing the presence or shape of the fruit in the orchard. In this paper, a 16-channel LIDAR (VLP-16) made of Velodyne was used to identify the shape of fruit trees. Solenoid valves were attached to the end parts of 12 nozzles of the orchard spraying machine for on/off control. The smart spraying machine implemented in this way was mounted on a vehicle capable of autonomous travel and performed selective control depending upon the shape of the fruit trees while traveling in the orchards. This is expected to significantly reduce the amounts of pesticides used in orchards and production costs.

• Journal of Platform Technology
• /
• v.9 no.4
• /
• pp.3-9
• /
• 2021

In this paper, a graph data model is introduced to effectively recognize the situation between each object on the road detected by vehicles or road infrastructure sensors. The proposed method builds a graph database by modeling each object on the road as a node of the graph and the relationship between objects as an edge of the graph, and updates object properties and edge properties in real time. In this case, the relationship between objects represented as edges is set when there is a possibility of approach between objects in consideration of the position, direction, and speed of each object. Finally, we propose a spatial indexing technique for graph nodes and edges to update the road object graph database represented through the proposed graph modeling method continuously in real time. To show the superiority of the proposed indexing technique, we compare the proposed indexing based database update method to the non-indexing update method through simulation. The results of the simulation show the proposed method outperforms more than 10 times to the non-indexing method.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.07a
• /
• pp.257-258
• /
• 2020

기존의 드론 조작은 초보자에게 어려웠다. 초보자의 경우 드론을 조종하다가 드론이 추락하거나 장애물에 걸려 프로펠러 등의 부품들이 손상되는 경우를 빈번하게 마주한다. 본 연구에서는 초보자 또한 드론 파손의 걱정 없이 드론의 조작을 더욱 쉽게 개선시키는 것을 전제로 뇌파와 보조입력인 음성인식을 이용한 드론 컨트롤러 기술을 적용하고자 한다. 현재 대중적으로 출시되어 있는 드론의 경우 호버링 기능을 포함시켜 드론의 추락 위험을 줄여주는 기능을 탑재하고 있다. 하지만 속도가 빠른 드론의 조작에 있어 미숙한 초보자들은 장애물과의 충돌 그리고 드론 착륙 시 기체손상 등의 위험에 대비하기 힘들다. 본 논문은 이러한 문제점들을 개선하기 위해 기존의 드론 컨트롤러 대신 특정한 동작을 상상할 때 발현되는 동작상상뇌파와 음성입력을 적용한 '동작상상뇌파와 음성인식을 이용한 드론 컨트롤러' 기술을 제안한다. 기존의 드론 컨트롤러와는 다르게 빅 데이터 처리기술인 머신러닝을 이용하여 뇌파 데이터를 처리하고 그 데이터들과 입력되는 뇌파 값을 비교하여 드론을 제어한다. 또한 뇌파의 발현이 안정적이지 못하는 상황을 대비한 보조입력인 음성인식을 이용하여 드론의 기체손상을 최소화 시킬 수 있다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.603-605
• /
• 2021

With the expansion of the indoor service-providing robot market and the electrification of automobiles, research on autonomous driving is being actively conducted. In general, in the case of outside, the location is mainly recognized through GPS, and location positioning is performed indoors using technologies such as WiFi, UWB (Ultra-Wide Band), VLP, LiDAR, and Vision. In this paper, we introduce a system for location-positioning using LED lights with different color temperatures in an indoor environment. After installing LED lights in a simulated environment such as a tunnel, it was shown that information about the current location can be obtained through the analysis of chromaticity values according to location. Through this, it is expected to be able to obtain information about the location of the vehicle in the tunnel and the movement of the device in a room such as a warehouse or a factory.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.239-241
• /
• 2022

Currently, in Korea, a multi-seater lane (HOV) and a designated lane system are being implemented to solve traffic congestion. However, in both systems, it is difficult to crack down on cases of violations without permission, so people are required to be assigned to areas that want to crack down. In this process, manpower and budget are inefficiently consumed. To compensate for these shortcomings, we propose the development of an integrated enforcement system through YOLO, a deep learning object recognition model. If the two systems are implemented and integrated using YOLO, they will have advantages in terms of manpower and budget over existing systems because only data learning and system maintenance are considered. In addition, in the case of violations in which it is difficult for the existing unmanned system to crack down, the effect of increasing the crackdown rate through continuous learning can be expected.

• Journal of Environmental Health Sciences
• /
• v.50 no.1
• /
• pp.54-65
• /
• 2024

Background: There is growing international recognition of the need for improvements to national chemical management systems for hazardous chemicals. The European Union has recently introduced the concept of 'essential uses' as a new approach to the management hazardous chemicals by limiting their uses. Objectives: This paper examines the concept of essential uses in chemical management and how to apply it through a case study of essential use. This approach is distinct from the current chemical management system, but seeks to improve its potential benefits by effectively restricting or gradually decreasing the use of hazardous substances. Methods: The concept of essential uses was introduced by reviewing the Montreal Protocol on Substances that Deplete the Ozone Layer, Cousins's three essentiality categories applied to PFAS, restriction options assessed in the PFAS restriction proposal under REACH, and the California Safer Consumer Products regulations prioritizing 6PPD in motor vehicle tires. Based on these essential uses concepts for PFAS and 6PPD, uses of benzene were classified in accordance with the essential uses approach for products using benzene in South Korea. Results: The essential use concept is able to manage the restriction and authorization of substances of concern through essential uses and non-essential uses and the feasible substitution of uses and substances. Conclusions: If the concept and methodology of essential uses are clearly established, they can be expected to shift the national chemical management paradigm from regulating substances to limiting uses under the existing substance management system.

• The Journal of Bigdata
• /
• v.9 no.1
• /
• pp.89-100
• /
• 2024

In the field of computer vision, models such as You Look Only Once (YOLO) and ResNet are widely used due to their real-time performance and high accuracy. However, to apply these models in real-world environments, factors such as runtime compatibility, memory usage, computing resources, and real-time conditions must be considered. This study compares the characteristics of three deep model runtimes: ONNX Runtime, TensorRT, and OpenCV DNN, and analyzes their performance on two models. The aim of this paper is to provide criteria for runtime selection for practical applications. The experiments compare runtimes based on the evaluation metrics of time, memory usage, and accuracy for vehicle license plate recognition and classification tasks. The experimental results show that ONNX Runtime excels in complex object detection performance, OpenCV DNN is suitable for environments with limited memory, and TensorRT offers superior execution speed for complex models.