• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.17 no.5
• /
• pp.173-187
• /
• 2018

This paper proposes a computer vision and deep learning-based technique for surveillance camera system for vehicle counting as one part of parking lot management system. We applied the You Only Look Once version 2 (YOLOv2) detector and come up with a deep convolutional neural network (CNN) based on YOLOv2 with a different architecture and two models. The effectiveness of the proposed architecture is illustrated using a publicly available Udacity's self-driving-car datasets. After training and testing, our proposed architecture with new models is able to obtain 64.30% mean average precision which is a better performance compare to the original architecture (YOLOv2) that achieved only 47.89% mean average precision on the detection of car, truck, and pedestrian.

• Proceedings of the KIEE Conference
• /
• 2015.07a
• /
• pp.1345-1346
• /
• 2015

본 논문에서는 보행자 및 배경 이미지로부터 HOG-PCA 특징을 추출하고 다항식 기반 RBFNNs(Radial Basis Function Neural Network) 패턴분류기과 최적화 알고리즘을 이용하여 보행자를 검출하는 시스템 설계를 제안한다. 입력 영상으로부터 보행자를 검출하기 위해 전처리 과정에서 HOG(Histogram of oriented gradient) 알고리즘을 통해 특징을 추출한다. 추출된 특징은 고차원이므로 패턴분류기 분류 시 많은 연산과 처리속도가 따른다. 이를 개선하고자 PCA (Principal Components Analysis)을 사용하여 저차원으로의 차원 축소한다. 본 논문에서 제안하는 분류기는 pRBFNNs 패턴분류기의 효율적인 학습을 위해 최적화 알고리즘인 PSO(Particle Swarm Optimization)을 사용하여 구조 및 파라미터를 최적화시켜 모델의 성능을 향상시킨다. 사용된 데이터로는 보행자 검출에 널리 사용되는 INRIA2005_person data set에서 보행자와 배경 영상을 각각 1200장을 학습 데이터, 검증 데이터로 구성하여 분류기를 설계하고 테스트 이미지를 설계된 최적의 분류기를 이용하여 보행자를 검출하고 검출률을 확인한다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.1249-1252
• /
• 2022

In this paper, we propose a conceptual framework that identifies video frames in motion containing the movement of people and vehicles in traffic videos. The automatic selection of video frames in motion is an important topic in security and surveillance video because the number of videos to be monitored simultaneously is simply too large due to limited human resources. The conventional method to identify the areas in motion is to compute the differences over consecutive video frames, which has been costly because of its high computational complexity. In this paper, we reduced the overall complexity by examining only the keyframes (or I-frames). The basic assumption is that the time period between I-frames is rather shorter (e.g., 1/10 ~ 3 secs) than the usual length of objects in motion in video (i.e., pedestrian walking, automobile passing, etc.). The proposed method estimates the possibility of videos containing motion between I-frames by evaluating the difference of consecutive I-frames with the long-time statistics of the previously decoded I-frames of the same video. The experimental results showed that the proposed method showed more than 80% accuracy in short surveillance videos obtained from different locations while keeping the computational complexity as low as 20 % compared to the HM decoder.

• PNF and Movement
• /
• v.22 no.1
• /
• pp.43-54
• /
• 2024

Purpose: Falls among persons older than 65 years are a significant concern due to their frequency and severity. This study aimed to develop a vest-type embedded artificial intelligence (AI) system capable of detecting and predicting falls in various scenarios. Methods: In this study, we established and developed a vest-type embedded AI system to judge and predict falls in various directions and situations. To train the AI, we collected data using acceleration and gyroscope values from a six-axis sensor attached to the seventh cervical and the second sacral vertebrae of the user, considering accurate motion analysis of the human body. The model was constructed using a neural network-based AI prediction algorithm to anticipate the direction of falls using the collected pedestrian data. Results: We focused on developing a lightweight and efficient fall prediction model for integration into an embedded AI algorithm system, ensuring real-time network optimization. Our results showed that the accuracy of fall occurrence and direction prediction using the trained fall prediction model was 89.0% and 78.8%, respectively. Furthermore, the fall occurrence and direction prediction accuracy of the model quantized for embedded porting was 87.0 % and 75.5 %, respectively. Conclusion: The developed fall detection and prediction system, designed as a vest-type with an embedded AI algorithm, offers the potential to provide real-time feedback to pedestrians in clinical settings and proactively prepare for accidents.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.25 no.1
• /
• pp.92-102
• /
• 2017

For an automotive surveillance radar system, fast-chirp train based FMCW (Frequency Modulated Continuous Wave) radar is a very effective method, because clutter and moving targets are easily separated in a 2D range-velocity map. However, pedestrians with low echo signals may be masked by strong clutter in actual field. To address this problem, we proposed in the previous work a clutter cancellation and moving target indication algorithm using the coherent phase method. In the present paper, we initially composed the test set-up using a 24 GHz FMCW transceiver and a real-time data logging board in order to verify this algorithm. Next, we created two indoor test environments consisting of moving human and stationary targets. It was found that pedestrians and strong clutter could be effectively separated when the proposed method is used. We also designed and implemented these algorithms in FPGA (Field Programmable Gate Array) in order to analyze the hardware and time complexities. The results demonstrated that the complexity overhead was nearly zero compared to when the typical method was used.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.1
• /
• pp.197-205
• /
• 2024

Development of AI and big data-based algorithms to advance and optimize the recognition and detection performance of various static/dynamic vehicles in front and around the vehicle at a time when securing driving safety is the most important point in the development and commercialization of autonomous vehicles. etc. are being studied. However, there are many research cases for recognizing the same vehicle by using the unique advantages of radar and camera, but deep learning image processing technology is not used, or only a short distance is detected as the same target due to radar performance problems. Therefore, there is a need for a convergence-based vehicle recognition method that configures a dataset that can be collected from radar equipment and camera equipment, calculates the error of the dataset, and recognizes it as the same target. In this paper, we aim to develop a technology that can link location information according to the installation location because data errors occur because it is judged as the same object depending on the installation location of the radar and CCTV (video).

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.5
• /
• pp.156-170
• /
• 2019

Rapid increasing of smartphones has changed people's lifestyles, and location-based services are providing a platform to provide various conveniences in accordance with these changes. In particular, it may provide convenience to many users if location-based services are provided in an indoor area such as subway station. However, it is still a difficult task to ensure accurate positioning result for guiding routes in subway stations. This study proposes a KAI-R system that allows all processes to be performed in one system for indoor navigation in subway stations. The proposed system includes a new pedestrian step detection method for continuous positioning along with an improved fusion positioning algorithm.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.13 no.1
• /
• pp.1-14
• /
• 2014

The expansion of the physical road in response to changes in social conditions and policy of the country has reached the limit. In order to alleviate congestion on the existing road to reconsider the effectiveness of this method should be asking. Currently, how to collect traffic information for management of the intersection is limited to point detection systems. Intelligent Transport Systems (ITS) was the traffic information collection system of point detection method such as through video and loop detector in the past. However, intelligent transportation systems of the next generation(C-ITS) has evolved rapidly in real time interval detection system of collecting various systems between the pedestrian, road, and car. Therefore, this study is designed to evaluate the development of an algorithm for queue length based real-time traffic signal control methodology. Four coordinates estimate on time-space diagram using the travel time each individual vehicle collected via the interval detector. Using the coordinate value estimated during the cycle for estimating the velocity of the shock wave the queue is created. Using the queue length is estimated, and determine the signal timing the total queue length is minimized at intersection. Therefore, in this study, it was confirmed that the calculation of the signal timing of the intersection queue is minimized.

• Spatial Information Research
• /
• v.21 no.3
• /
• pp.11-19
• /
• 2013

Compared with traditional video monitoring systems that provide a video-recording function as a main service, an intelligent video monitoring system is capable of extracting/tracking objects and detecting events such as car accidents, traffic congestion, pedestrian detection, and so on. Thus, the object tracking is an essential function for various intelligent video monitoring and surveillance systems. In this paper, we propose a background and local histogram-based object tracking approach for intelligent video monitoring systems. For robust object tracking in a live situation, the result of optical flow and local histogram verification are combined with the result of background subtraction. In the proposed approach, local histogram verification allows the system to track target objects more reliably when the local histogram of LK position is not similar to the previous histogram. Experimental results are provided to show the proposed tracking algorithm is robust in object occlusion and scale change situation.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.12
• /
• pp.561-568
• /
• 2021

In recent years, autonomous driving technologies have become a high-value-added technology that attracts attention in the fields of science and industry. For smooth Self-driving, it is necessary to accurately detect an object and estimate its movement speed in real time. CNN-based deep learning algorithms and conventional dense optical flows have a large consumption time, making it difficult to detect objects and estimate its movement speed in real time. In this paper, using a single camera image, fast object detection was performed using the YOLOv5 algorithm, a deep learning algorithm, and fast estimation of the speed of the object was performed by using a local dense optical flow modified from the existing dense optical flow based on the detected object. Based on this algorithm, we present a system that can predict the collision time and probability, and through this system, we intend to contribute to prevent tram accidents.