• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.6
• /
• pp.1329-1338
• /
• 2021

There are rarely effective methods to help visually impaired people when they want to know the expiration date of products excepted to only Braille. In this study, we developed an expiration date notification system based on YOLO and OCR for visually impaired people. The handicapped people can automatically know the expiration date of a specific product by using our system without the help of a caregiver, fast and accurately. The proposed system is worked by four different steps: (1) identification of a target product by scanning its barcode; (2) segmentation of an image area with the expiration date using YOLO; (3) classification of the expiration date by OCR: (4) notification of the expiration date by TTS. Our system showed an average classification accuracy of about 86.00% when blindfolded subjects used the proposed system in real-time. This result validates that the proposed system can be potentially used for visually impaired people.

• Journal of Platform Technology
• /
• v.9 no.4
• /
• pp.90-96
• /
• 2021

Most of the vehicles currently under development and in operation are equipped with various IoT sensors, but some of the factors that cause car accidents are relatively difficult to detect. One of the major risk factors among these factors is black ice. Black ice is one of the factors most likely to cause major accidents, as it can affect all vehicles passing through areas covered with black ice. Therefore, black ice detection technique is essential to prevent major accidents. For this purpose, some studies have been carried out in the past, but unrealistic factors have been reflected in some parts, so research to supplement this is needed. In this paper, we tried to detect black ice by analyzing color images using the CNN technique, and we succeeded in detecting black ice to a certain level. However, there were differences from previous studies, and the reason was analyzed.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.448-452
• /
• 2006

본 논문에서는 특정 마커를 사용하는 실감형 증강현실 시스템 상에서 카메라가 비정형적인 움직임을 하는 경우에 대하여, 다중 마커를 사용한 떨림 현상을 줄인 실시간 움직임 추적 기법을 제안하고자 한다. 카메라의 움직임을 추정하기 위하여 카메라와 마커 사이의 변환을 계산해야 한다. 이미지로부터 검출된 각 마커의 네 모서리 점들을 이용하여, 각 마커에 대한 변환을 계산한다. 마커는 서로 다른 로컬 좌표계를 가지고 있고, 마커에 대한 변환은 해당 마커의 좌표계에 의해 정의된다. 다중 마커의 로컬 좌표계로부터 최적의 카메라 움직임을 추정하기 위한 정합 알고리즘을 제안한다. 정합을 위한 방법으로 레퍼런스 마커를 사용한다. 레퍼런스 마커는 정합 과정에서 자동적으로 선택된다. 레퍼런스 마커를 기준으로 각 마커의 변환에 대해 신뢰성(confidence rate)을 기반으로 가중치를 적용함으로써 최적의 카메라 움직임을 추정할 수 있다. 또한 추정된 카메라의 움직임의 최적화를 위하여 히스토리 버퍼를 사용하여 떨림 현상을 제거하는 방법을 제안한다. 추정된 카메라의 위치에 대한 평균 필터 및 중간 필터의 개념과 유사한 보정 방법을 통해 떨림 현상을 제거한다. 실험을 통해 다른 방법들과 비교한 우리가 제안한 방법의 정확성을 확인할 수 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.766-768
• /
• 2005

최근 대형 디스플레이 및 웨어러블 컴퓨터의 등장과 함께 키보드와 마우스를 사용하는 일반 데스크탑 환경에서 벗어난 컴퓨터와의 자연스러운 상호 작용 연구가 활발히 진행되고 있다. 본 논문에서는 스크린 정면에 놓인 프로젝터가 스크린과 그 위에 놓인 사용자의 손 위에 화면을 투영할 때 PC 카메라로 입력된 프레임 속에서 손의 영역을 인식하여 컴퓨터와 상호작용하게 하고자 한다. 이 경우에 투영된 빛이 사용자의 손 위에도 합쳐짐으로 인하여 피부의 고유색이 사라진다. 또한, 투영되는 화면이 사용자와 컴퓨터의 상호 작용에 따라 추정할 수 없이 변함에 따른 적응적 인식 방법이 필요하다. 따라서, 본 논문에서는 손을 인식하기에 앞서 스크린에 투영될 원본 이미지에 대해 칼라 보정을 수행하여 추정되는 카메라 입력 프레임을 생성한다. 이를 위해 우선 백색 영상을 투영하여 프레임 내의 자기 오차 맵을 생성한 후 R,G,B 채널 별로 원본 값에 대한 카메라 반응 값들을 룩업 테이블에 저장한다. 이를 통해 원본 이미지에 대해 칼라보정을 수행하고, 생성된 카메라 추정 프레임과 실제 카메라로 입력된 프레임 간 자기 성분을 비교하여 손 영역을 검출한다. 실험 결과, 주변의 조명 상태나 프로젝터 및 카메라의 위치에 관계없이 안정적인 인식 결과를 보였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.74-77
• /
• 2021

최근 딥러닝을 이용하여 객체 인식 학습을 위한 데이터셋을 구축하는데 있어 시간과 인력을 단축하기 위해 인조 데이터를 생성하는 연구가 진행되고 있다. 하지만 실제 환경과 관계없이 임의의 배경에 배치되어 구축된 데이터셋으로 학습된 네트워크를 실제 환경으로 구성된 데이터셋으로 테스트할 경우 인식률이 저조하다. 이에 본 논문에서는 실제 배경 이미지에 객체 이미지를 합성하고, 다양성을 위해 3차원으로 회전하여 증강하는 인조 데이터셋 생성 시스템을 제안한다. 제안된 방법으로 구축된 인조 데이터셋으로 학습한 네트워크와 실제 데이터셋으로 학습된 네트워크의 인식률을 비교한 결과, 인조 데이터셋의 성능이 실제 데이터셋의 성능보다 2% 낮았지만, 인조 데이터셋을 구축하는 시간이 실제 데이터셋을 구축하는 시간보다 약 11배 빨라 시간적으로 효율적인 데이터셋 구축 시스템임을 증명하였다.

• Journal of KIISE:Software and Applications
• /
• v.31 no.3
• /
• pp.379-386
• /
• 2004

Video segmentation is a fundamental task in video indexing and it includes two kinds of shot change detections such as the abrupt transition and the gradual transition. The abrupt shot boundaries are detected by computing the image-based distance between adjacent frames and comparing this distance with a pre-determined threshold value. However, the gradual shot boundaries are difficult to detect with this approach. To overcome this difficulty, we propose the method that detects gradual transition in the MPEG compressed video using the HMM (Hidden Markov Model). We take two different HMMs such as a discrete HMM and a continuous HMM with a Gaussian mixture model. As image features for HMM's observations, we use two distinct features such as the difference of histogram of DC images between two adjacent frames and the difference of each individual macroblock's deviations at the corresponding macroblock's between two adjacent frames, where deviation means an arithmetic difference of each macroblock's DC value from the mean of DC values in the given frame. Furthermore, we obtain the DC sequences of P and B frame by the first order approximation for a fast and effective computation. Experiment results show that we obtain the best detection and classification performance of gradual transitions when a continuous HMM with one Gaussian model is taken and two image features are used together.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.3
• /
• pp.389-395
• /
• 2022

Because of improvement of deep learning techniques, deep learning using computer vision such as classification, detection and segmentation has also been used widely at many fields. Expecially, automatic driving is one of the major fields that applies computer vision systems. Also there are a lot of works and researches to combine multiple tasks in a single network. In this study, we propose the network that predicts the individual depth of pedestrians and vehicles. Proposed model is constructed based on YOLOv3 for object detection and Monodepth for depth estimation, and it process object detection and depth estimation consequently using encoder and decoder based on hard parameter sharing. We also used attention module to improve the accuracy of both object detection and depth estimation. Depth is predicted with monocular image, and is trained using self-supervised training method.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.4
• /
• pp.715-722
• /
• 2022

If you don't regularly go to the dentist to check your teeth, it is difficult to notice cavities or various diseases of your teeth until you have pain or discomfort. Dental plaque is produced by the combination of food or foreign substances and bacteria in the mouth. Starch breaks down from the bacteria that form tartar. The acid that occurs at this time melts the enamel of the teeth and becomes a cavity. So tartar management is important. Poppyrin, the metabolism of bacteria in the mouth, reacts at 405 nm wavelengths and becomes red fluorescent, which can be seen by imaging through certain wavelength filters. By the above method, Frag and tartar are fluorescently detected and photographed with a yellow series of filters that pass wavelengths of 500 nm or more. It uses MATLAB to detect and display red fluorescence through image processing. Using the difference in voltage between normal teeth and tartar through an optical measuring circuit, it was connected to an Arduino and displayed on the LCD. This allows the user to know the presence and location of dental plaque more accurately.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.13 no.5
• /
• pp.19-28
• /
• 2017

Various methods for reducing the load on the server have been implemented in performing face recognition remotely by the spread of IP security cameras. In this paper, IP surveillance cameras at remote sites are input through a DSP board equipped with face detection function, and then face detection is performed. Then, the facial region image is transmitted to the server, and the face recognition processing is performed through face recognition distributed processing. As a result, the overall server system load and significantly reduce processing and real-time face recognition has the advantage that you can perform while linked up to 256 cameras. The technology that can accomplish this is to perform 64-channel face recognition per server using distributed processing server technology and to process face search results through 250 camera channels when operating four distributed processing servers there was.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.10
• /
• pp.35-42
• /
• 2020

In this study, we propose a crack detection method using limited data with a U-Net based image inpainting technique that is a modified unsupervised anomaly detection method. Concrete cracking occurs due to a variety of causes and is a factor that can cause serious damage to the structure in the long term. In general, crack investigation uses an inspector's visual inspection on the concrete surfaces, which is less objective in judgment and has a high possibility of human error. Therefore, a method with objective and accurate image analysis processing is required. In recent years, the methods using deep learning have been studied to detect cracks quickly and accurately. However, when the amount of crack data on the building or infrastructure to be inspected is small, existing crack detection models using it often show a limited performance. Therefore, in this study, an unsupervised anomaly detection method was used to augment the data on the object to be inspected, and as a result of learning using the data, we confirmed the performance of 98.78% of accuracy and 82.67% of harmonic average (F1_Score).