• 한국HCI학회:학술대회논문집
• /
• 한국HCI학회 2009년도 학술대회
• /
• pp.547-551
• /
• 2009

최근, 필름복원은 다양한 멀티미디어 출현과 영상보존의 중요성으로 많은 연구자들로부터 관심을 받고 있다. 블로치(blotch)는 오래된 영상에서 나타나는 대표적인 손상요인이다. 따라서 본 논문에서는 객체의 컬러특징과 방향분포 변화를 이용한 블로치 검출방법을 제안한다. 제안된 방법은 두가지 모듈로 구성 된다. 블로치의 불연속적인 특징을 이용한 SROD 검출기로 불로치의 후보지를 검출하고, 후보지로부터 블로치의 컬러와 형태 특징을 이용한 신경망으로 블로치 영역을 검출한다. 제안된 방법을 평가 하기 위해 실제 오래된 영상으로부터 실험 하였다.

• Journal of Information Display
• /
• 제13권1호
• /
• pp.7-16
• /
• 2012

In this paper, the problem concerning the color accuracy of imaging systems using color filters is examined. It is shown that the only solution to the problem is to build systems with the spectral response matching the CIE curves as closely as possible. If the spectral response does not closely match the CIE curves, it was demonstrated that calibration cannot solve the problem and will result in very unstable colorimeters. A practical solution that uses telecentric lenses on the sensor side in addition to dedicated color filters for each CCD detector is presented. For systems that closely match the CIE curves, an innovative method of improving the color accuracy based on the precise measurement of the spectral response is presented. The small discrepancies in the spectral response with regard to the CIE curves are corrected in different ways during the measurements. Finally, it is shown that the tristimulus calibration that is used for display measurement is very unstable for systems without CIE matching and is much more stable with systems that closely match the CIE curves.

• 방송공학회논문지
• /
• 제23권3호
• /
• pp.361-368
• /
• 2018

자율주행 자동차의 연구가 빠르게 발전하는 가운데 보행자 검출에 대한 연구 또한 성공적으로 진행되고 있다. 그러나 대부분의 연구에서 사용되는 데이터셋이 컬러영상을 기반하고 있고 또한 보행자의 인식이 상대적으로 쉬운 영상이 많다. 컬러 영상의 경우 보행자가 빛에 노출되는 정도에 따라 영상에 제대로 포착이 되지 않을 수 있고 이로 인해 기존 방식들로는 이러한 보행자를 제대로 검출하지 못하는 상황이 발생한다. 따라서 본 논문에서는 DNN (deep neural network) 기반 컬러 영상과 열 영상을 이용한 다중 스펙트럼 보행자 검출 기법을 제안하고자 한다. 기존의 SSD (single shot multibox detector) 기법을 기반으로 하여 컬러 영상과 열 영상을 동시에 활용하는 퓨전 네트워크 구조를 제안한다. 실험은 KAIST의 데이터셋을 이용하여 실시하였고 제안한 기법인 SSD-H (SSD-Halfway fusion)의 방식이 KAIST 보행자 검출기준의 기준치보다 18.18% 낮은 miss rate를 획득하였고 또한 기존 halfway fusion 기법에 비해 최소 2.1% 낮은 miss rate를 획득하였다.

• 한국분무공학회지
• /
• 제20권4호
• /
• pp.230-235
• /
• 2015

The demands for reliable particle cloud temperature measurement exist in many process industries and scientific researches. Particle cloud temperature measurements depend on radiation thermometry at two or more color bands. In this study, we developed a sensitive, fast response and compact online infrared two-color probe to measure the temperature of a particle cloud in a phase of two field flow (solid-gas). The probe employs a detector contained two InGaAs photodiodes with different spectral responses in the same optical path, which allowed a compact probe design. The probe was designed to suit temperature measurements in harsh environments with the advantage of durability. The developed two-color probe is capable of detecting particle cloud temperature as low as $300^{\circ}C$, under dynamic conditions.

• 한국멀티미디어학회논문지
• /
• 제18권10호
• /
• pp.1149-1156
• /
• 2015

This paper proposes the hand contour detector which is robust to noises. Existing methods reduce noises by applying morphology to extracted edges, detect finger tips by using the center of hands, or exploit the intersection of curves from hand area candidates based on J-value segmentation(JSEG). However, these approaches are so vulnerable to noises that are prone to detect non-hand parts. We propose the noise tolerant hand contour detection method in which non-skin area noises are removed by applying skin area detection, contour detection, and a threshold value. By using the implemented system, we observed that the system was successfully able to detect hand contours.

• 한국광학회:학술대회논문집
• /
• 한국광학회 2000년도 제11회 정기총회 및 00년 동계학술발표회 논문집
• /
• pp.240-241
• /
• 2000

Sapphire fiber has been used to provide an optical path for the total radiation pyrometry. In measuring the temperature, we use the two-color detector, which consists of a high-performance Silicon detector mounted in a "sandwich" configuration over a Germanium detector. Sapphire fiber can withstand high temperature in the molten steel for two and a half hours. The maximum value of the error is the $\pm$2.5$^{\circ}C$ in the range of 152$0^{\circ}C$~1$600^{\circ}C$. This paper presents the simple scheme for measuring the molten steel temperature in the blast furnace of the iron ＆ steel making process.g process.

• 한국화재소방학회논문지
• /
• 제31권2호
• /
• pp.44-51
• /
• 2017

성능위주 소방설계(PBD)의 과정에서 요구피난시간(RSET) 산정의 신뢰성을 확보하기 위해서는 화재시뮬레이션을 이용한 정확한 연기감지기 작동시간 예측이 필수적이다. 본 연구의 목적은 FDS 기반의 연기감지기 수치모델에서 요구되는 입력인자의 정확도를 개선시키는 것이다. 이를 위하여 선행연구에서 적용된 화재감지기 시험장치(FDE)의 개선이 이루어졌다. 구체적으로 FDE 내부의 유동 및 연기농도 균일성이 개선되었으며, 연기입자의 전방산란 저감을 통해 감지기가 작동되는 순간의 정확한 광 소멸률이 측정되었다. 개선된 FDE를 이용한 입력인자는 기존 결과와 정량적으로 상당한 변화를 보이고 있으며, 이온화식 감지기에 비해 광전식 감지기에서 더 큰 차이가 확인되었다. 연기감지기의 작동조건은 감지기 종류, 가연물, 연기입자 및 색상에 따라 큰 차이가 발생됨을 고려할 때, PBD의 신뢰성을 향상시키기 위하여 향후 연구에서는 보다 다양한 감지기 및 가연물에 대한 입력인자 DB가 구축되어야 할 것이다.

• 한국멀티미디어학회논문지
• /
• 제18권5호
• /
• pp.610-619
• /
• 2015

*In this paper, we propose a real-time face tracking system using an adaptive face detector and a tracking algorithm. An image is divided into the regions of background and face candidate by a real-time updated skin color identifying system in order to accurately detect facial features. The facial characteristics are extracted using the five types of simple Haar-like features. The extracted features are reinterpreted by Principal Component Analysis (PCA), and the interpreted principal components are processed by Support Vector Machine (SVM) that classifies into facial and non-facial areas. The movement of the face is traced by Kalman filter and Mean shift, which use the static information of the detected faces and the differences between previous and current frames. The proposed system identifies the initial skin color and updates it through a real-time color detecting system. A similar background color can be removed by updating the skin color. Also, the performance increases up to 20% when the background color is reduced in comparison to extracting features from the entire region. The increased detection rate and speed are acquired by the usage of Kalman filter and Mean shift.

• 제어로봇시스템학회논문지
• /
• 제20권9호
• /
• pp.930-935
• /
• 2014

A Gabor filter is a linear filter used for edge detectionas frequency and orientation representations of Gabor filters are similar to those of the human visual system. In this thesis, we propose a pedestrian detection algorithm using a Gabor filter bank. In order to extract the features of the pedestrian, we use various image processing algorithms and data structure algorithms. First, color image segmentation is performed to consider the information of the RGB color space. Second, histogram equalization is performed to enhance the brightness of the input images. Third, convolution is performed between a Gabor filter bank and the enhanced images. Fourth, statistical values are calculated by using the integral image (summed area table) method. The calculated statistical values are used for the feature matrix of the pedestrian area. To evaluate the proposed algorithm, the INRIA pedestrian database and SVM (Support Vector Machine) are used, and we compare the proposed algorithm and the HOG (Histogram of Oriented Gradient) pedestrian detector, presentlyreferred to as the methodology of pedestrian detection algorithm. The experimental results show that the proposed algorithm is more accurate compared to the HOG pedestrian detector.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권12호
• /
• pp.2355-2373
• /
• 2011

Vision-based vehicle detector systems are becoming increasingly important in ITS applications. Real-time operation, robustness, precision, accurate estimation of traffic parameters, and ease of setup are important features to be considered in developing such systems. Further, accurate vehicle detection is difficult in varied complex traffic environments. These environments include changes in weather as well as challenging traffic conditions, such as shadow effects and jams. To meet real-time requirements, the proposed system first applies a color background to extract moving objects, which are then tracked by considering their relative distances and directions. To achieve robustness and precision, the color background is regularly updated by the proposed algorithm to overcome luminance variations. This paper also proposes a scheme of feedback compensation to resolve background convergence errors, which occur when vehicles temporarily park on the roadside while the background image is being converged. Next, vehicle occlusion is resolved using the proposed prior split approach and through reasoning for rule-based tracking. This approach can automatically detect straight lanes. Following this step, trajectories are applied to derive traffic parameters; finally, to facilitate easy setup, we propose a means to automate the setting of the system parameters. Experimental results show that the system can operate well under various complex traffic conditions in real time.