• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2017년도 추계학술발표대회
• /
• pp.134-137
• /
• 2017

Follow-me function of drones is new and attractive for selfie drone users, where the drone autonomously follows and capture the user. Currently the products use the difference between GPS's in the drone and user side mobile GCS, but the targeting accuracy is not satisfactory owing to the low accuracy of GPS data, often the order of ten meters. We designed a new follow-me mode algorithm that utilizes the accuracy of visual tracking algorithm and the reliability of GPS-based. The experiment shows that proposed follow-me can capture much accurately the target user in the center of video content than GPS-only methods, and recover the vision algorithm failure quickly in 5-10 seconds.

• IEIE Transactions on Smart Processing and Computing
• /
• 제4권4호
• /
• pp.281-290
• /
• 2015

Human action recognition from a video scene has remained a challenging problem in the area of computer vision and pattern recognition. The development of the low-cost RGB depth camera (RGB-D) allows new opportunities to solve the problem of human action recognition. In this paper, we present a comprehensive review of recent approaches to human action recognition based on depth maps, skeleton joints, and other hybrid approaches. In particular, we focus on the advantages and limitations of the existing approaches and on future directions.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 1999년도 KOBA 방송기술 워크샵 KOBA Broadcasting Technology Workshop
• /
• pp.171-176
• /
• 1999

Template matching is one of the most often used techniques for machine vision applications to find a template of size M$\times$M or subimage in a scene image of size N$\times$N. Most template matching methods, however, require pixel operations between the template and the image under analysis resulting in high computational cost of O(M2N2). So in this thesis, we present a two stage template matching method. In the first stage, we use a novel low cost feature whose complexity is approaching O(N2) to select matching candidates. In the second stage, we use conventional template matching method to find out the exact matching point. We compare the result with other methods in terms of complexity, efficiency and performance. Proposed method was proved to have constant time complexity and to be quite invariant to noise.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2014년도 추계학술발표대회
• /
• pp.947-950
• /
• 2014

CCTV와 차량용 블랙박스 등의 영상기반 감시장비들로 사회안전망이 구축되고 있다. 하지만 디지털 영상 획득 센서는 실세계의 다이나믹 레인지를 온전히 감지하지 못한다는 한계점을 가지고 있는데 이로 인해 역광과 같은 특정 조명 조건하에서는 발생하는 움직임들을 감지하지 못하는 문제가 있다. 이러한 문제점을 해결하기 위해 종래에는 HDR 이미지를 사용하는데, 움직임이 많은 영상에 적용하기 어렵다. 별도의 WDR 이미지 센서를 사용할 수도 있으나 가격이 비싸고 영상처리가 복잡하다는 단점이 있다. 따라서, 본 논문에서는 프레임을 목표 다이내믹 레인지별로 그룹핑하고 프레임 그룹별로 노출시간을 달리하는 다중노출 방식을 제안한다. 이 방식에 따르면 어떤 조명 조건 상황에서도 물체의 변화를 모두 검출할 수 있으며 기존 이미지 센서와 영상 감지 시스템을 그대로 사용하기에 저비용으로 구현이 가능하다는 장점이 있다.

• 한국소음진동공학회논문집
• /
• 제18권12호
• /
• pp.1229-1236
• /
• 2008

A set of modal parameters of a stay-cable have been extracted fi:on a moving picture captured by a digital camera supported by shaking hands. It is hard to identify the center of targets attached on the cable surface from the blurred cable motion image, because of the high speed motion of cable, low sampling frequency of camera, and the shaking effect of camera. This study proposes a multi-template matching algorithm to resolve such difficulties. In addition, a sensitivity-based system identification algorithm is introduced to extract the natural frequencies and damping ratios from the ambient cable vibration data. Three sets of vibration tests are conducted to examine the validity of the proposed algorithms. The results show that the proposed technique is pretty feasible for extracting modal parameters from the severely shaking motion pictures.

• 반도체디스플레이기술학회지
• /
• 제17권1호
• /
• pp.84-87
• /
• 2018

As the propeller end speed increases, the propeller surface is damaged in the process of bubble formation and dropout. It is intended to prevent the corrosion of the propeller by modifying the shape of the end through the winglet structure to mitigate the cavitation phenomenon. In the case of conventional SUS materials, the cost of production is so high that plastic materials are used to prevent corrosion. This paper aims to mitigate the cavitation by deforming the shape of the end through the winglet structure by using the SMC composite material of the propeller using the existing SUS.

• 한국멀티미디어학회논문지
• /
• 제22권4호
• /
• pp.432-442
• /
• 2019

Most current human action recognition methods based on deep learning methods. It is required, however, a very high computational cost. In this paper, we propose an action change detection method to reduce repetitive human action recognition tasks. In reality, simple actions are often repeated and it is time consuming process to apply high cost action recognition methods on repeated actions. The proposed method decides whether action has changed. The action recognition is executed only when it has detected action change. The action change detection process is as follows. First, extract the number of non-zero pixel from motion history image and generate one-dimensional time-series data. Second, detecting action change by comparison of difference between current time trend and local extremum of time-series data and threshold. Experiments on the proposed method achieved 89% balanced accuracy on action change data and 61% reduced action recognition repetition.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권1호
• /
• pp.166-179
• /
• 2021

Recently, with the development of technology, computer vision research based on the human visual system has been actively conducted. Saliency maps have been used to highlight areas that are visually interesting within the image, but they can suffer from low performance due to external factors, such as an indistinct background or light source. In this study, existing color, brightness, and contrast feature maps are subjected to multiple shape and orientation filters and then connected to a fully connected layer to determine pixel intensities within the image based on location-based weights. The proposed method demonstrates better performance in separating the background from the area of interest in terms of color and brightness in the presence of external elements and noise. Location-based weight normalization is also effective in removing pixels with high intensity that are outside of the image or in non-interest regions. Our proposed method also demonstrates that multi-filter normalization can be processed faster using parallel processing.

• 센서학회지
• /
• 제30권1호
• /
• pp.15-19
• /
• 2021

Image processing is one of the major techniques that are used for computer vision. Nowadays, researchers are using machine learning and deep learning for the aforementioned task. In recent years, digit recognition tasks, i.e., automatic meter recognition approach using electric or water meters, have been studied several times. However, two major issues arise when we talk about previous studies: first, the use of the deep learning technique, which includes a large number of parameters that increase the computational cost and consume more power; and second, recent studies are limited to the detection of digits and not storing or providing detected digits to a database or mobile applications. This paper proposes a system that can detect the digital number of meter readings using a lightweight deep neural network (DNN) for low power consumption and send those digits to an Android mobile application in real-time to store them and make life easy. The proposed lightweight DNN is computationally inexpensive and exhibits accuracy similar to those of conventional DNNs.

• 전자통신동향분석
• /
• 제36권4호
• /
• pp.1-12
• /
• 2021

The competition to lead the next generation of mobile technologies, 6G, is underway while the deployment of 5G has not been implemented worldwide. ITU-R plans to develop technical requirements and standards after completing the 6G Vision by 2023. It can be considered too early to have a concrete view of the 6G core network architecture from this timeline. However, major stakeholders have started making their presence felt by publishing their views. From updated analysis on the technology and service trends proposed, we present a list of research directions on 6G core network from several perspectives: distribution of network functions to nearer edge locations; future fixed-mobile convergence, including low earth orbit satellites; highly-precise QoS guarantee; supporting an extremely wide variety of service requirements; AI-native automation and intelligence; and aligning with the evolution of radio access network.