• 로봇학회논문지
• /
• 제14권1호
• /
• pp.14-21
• /
• 2019

In the ground environment, mobile robot research uses sensors such as GPS and optical cameras to localize surrounding landmarks and to estimate the position of the robot. However, an underwater environment restricts the use of sensors such as optical cameras and GPS. Also, unlike the ground environment, it is difficult to make a continuous observation of landmarks for location estimation. So, in underwater research, artificial markers are installed to generate a strong and lasting landmark. When artificial markers are acquired with an underwater sonar sensor, different types of noise are caused in the underwater sonar image. This noise is one of the factors that reduces object detection performance. This paper aims to improve object detection performance through distortion and rotation augmentation of training data. Object detection is detected using a Faster R-CNN.

• Journal of Astronomy and Space Sciences
• /
• 제33권4호
• /
• pp.313-318
• /
• 2016

Communication, Ocean And Meteorological Satellite (COMS) Meteorological Imager (MI) images are processed for radiometric and geometric correction from raw image data. When intermediate image data are matched and compared with reference landmark images in the geometrical correction process, various techniques for edge detection can be applied. It is essential to have a precise and correct edged image in this process, since its matching with the reference is directly related to the accuracy of the ground station output images. An edge detection method based on neural networks is applied for the ground processing of MI images for obtaining sharp edges in the correct positions. The simulation results are analyzed and characterized by comparing them with the results of conventional methods, such as Sobel and Canny filters.

• Imaging Science in Dentistry
• /
• 제51권3호
• /
• pp.299-306
• /
• 2021

Purpose: This study aimed to propose a fully automatic landmark identification model based on a deep learning algorithm using real clinical data and to verify its accuracy considering inter-examiner variability. Materials and Methods: In total, 950 lateral cephalometric images from Yonsei Dental Hospital were used. Two calibrated examiners manually identified the 13 most important landmarks to set as references. The proposed deep learning model has a 2-step structure-a region of interest machine and a detection machine-each consisting of 8 convolution layers, 5 pooling layers, and 2 fully connected layers. The distance errors of detection between 2 examiners were used as a clinically acceptable range for performance evaluation. Results: The 13 landmarks were automatically detected using the proposed model. Inter-examiner agreement for all landmarks indicated excellent reliability based on the 95% confidence interval. The average clinically acceptable range for all 13 landmarks was 1.24 mm. The mean radial error between the reference values assigned by 1 expert and the proposed model was 1.84 mm, exhibiting a successful detection rate of 36.1%. The A-point, the incisal tip of the maxillary and mandibular incisors, and ANS showed lower mean radial error than the calibrated expert variability. Conclusion: This experiment demonstrated that the proposed deep learning model can perform fully automatic identification of cephalometric landmarks and achieve better results than examiners for some landmarks. It is meaningful to consider between-examiner variability for clinical applicability when evaluating the performance of deep learning methods in cephalometric landmark identification.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.677-680
• /
• 2003

Perceptual landmarks are an effective solution for a mobile robot realizing steady and reliable long distance navigation. But the prerequisite is those landmarks must be detected and recognized robustly at a higher speed under various lighting conditions. This made image processing more complicated so that its speed and reliability can not be both satisfied at the same time. Color based target detection technique can separate target color regions from non-target color regions in an image with a faster speed, and better results were obtained only under good lighting conditions. Moreover, in the case that there are other things with a target color, we have to consider other target features to tell apart the target from them. Such thing always happens when we detect a target with its single character. On the other hand, we can generally search for only one target for each time so that we can not make use of landmarks efficiently, especially when we want to make more landmarks work together. In this paper, by making use of the moment invariants of each landmark, we can not only search specified target from separated color region but also find multi-target at the same time if necessary. This made the finite landmarks carry on more functions. Because moment invariants were easily used with some low level image processing techniques, such as color based target detection and gradient runs based target detection etc, and moment invariants are more reliable features of each target, the ratio of target detection were improved. Some necessary experiments were carried on to verify its robustness and efficiency of this method.

• 한국측량학회지
• /
• 제35권5호
• /
• pp.339-348
• /
• 2017

Due to recent improvements in computer processing speed and image processing technology, researches are being actively carried out to combine information from a camera with existing GNSS (Global Navigation Satellite System) and dead reckoning. In this study, the mathematical model based on SPR (Single Photo Resection) is derived for image-based assistant algorithm for vehicle positioning. Simulation test is performed to analyze factors affecting SPR. In addition, GNSS/on-board vehicle sensor/image based positioning algorithm is developed by combining image-based positioning algorithm with existing positioning algorithm. The performance of the integrated algorithm is evaluated by the actual driving test and landmark's position data, which is required to perform SPR, based on simulation. The precision of the horizontal position error is 1.79m in the case of the existing positioning algorithm, and that of the integrated positioning algorithm is 0.12m at the points where SPR is performed. In future research, it is necessary to develop an optimized algorithm based on the actual landmark's position data.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2007년도 춘계학술대회 학술발표 논문집 제17권 제1호
• /
• pp.143-146
• /
• 2007

객체의 특징점을 추출할 때, 일반적으로 모델 기반 접근을 사용한다. 본 논문에서는 이러한 모델 기반 특징점 추출 알고리즘으로 PCA를 근간으로 하는 Active Appearance Model을 이용하는데, 기존의 AAM 알고리즘은 모든 특징점을 하나의 군집으로 기준하여 PCA를 수행하지만 본 논문에서는 이것을 각 주요 부위별 학습 모델로 분리하여 수행한다. 그리고 이러한 모델에서 특징점을 찾을 때, 발산하는 문제에 빠지지 않기 위한 방법을 제시한다. 제시한 방법의 모델을 이용하여 실험 할 경우의 결과와 이를 통한 개별 모델의 특성에 대하여 파악한 결과를 제시한다.

• Journal of Multimedia Information System
• /
• 제5권1호
• /
• pp.15-20
• /
• 2018

There are a lot of problems in the face detection area. One of them is detecting faces by facial features and reducing number of the false negatives and positions. This paper is directed to solve this problem by the proposed triangle method. Also, this paper explans cascades, Haar-like features, AdaBoost, HOG. We propose a scheme using 12-net, 24-net, 48-net to scan images and improve efficiency. Using triangle method for frontal pose, B and B1 methods for other poses in neural networks are proposed.

• Ocean Systems Engineering
• /
• 제10권4호
• /
• pp.435-449
• /
• 2020

This paper proposes a method to estimate the underwater target object's yaw angle using a sonar image. A simulator modeling imaging mechanism of a sonar sensor and a generative adversarial network for style transfer generates realistic template images of the target object by predicting shapes according to the viewing angles. Then, the target object's yaw angle can be estimated by comparing the template images and a shape taken in real sonar images. We verified the proposed method by conducting water tank experiments. The proposed method was also applied to AUV in field experiments. The proposed method, which provides bearing information between underwater objects and the sonar sensor, can be applied to algorithms such as underwater localization or multi-view-based underwater object recognition.

• 한국통신학회논문지
• /
• 제31권12C호
• /
• pp.1157-1164
• /
• 2006

본 논문에서는 마이크로 BGA의 패키지와 볼의 정확한 위치와 사이즈를 측정하기 위한 방법을 제안하였다. 정확하게 BGA의 결함을 찾아내기 위해, 패키지와 볼의 위치를 찾아내는데 중점을 두었다. 라벨링한 후, 특징 파라미터를 이용하여 패키지와 볼 성분만을 검출하였다. 패키지 부분을 검출한 후, 패키지에 대한 정보를 입력 파라미터로 사용하여 사각형 모델로 패키지의 사이즈를 측정하였다. 또한 볼 부분을 검출한 후, 볼 부분에 대한 정보를 입력 파라미터로 사용하여 원형 모델로 볼의 위치와 지름을 측정하였다. 실제 길이를 측정하기 위하여 landmark에 근거한 calibration을 수행하였으며 SEM으로 볼을 측정한 데이터를 기준으로 측정치와 비교하였다. 위의 실험으로부터 제안 기법에 의한 볼의 반지름 측정값의 정확도가 평균 94%가 되는 사실을 확인하였다.

• 한국멀티미디어학회논문지
• /
• 제23권12호
• /
• pp.1454-1463
• /
• 2020

An important factor in eye-tracking PC interface for general paralyzed patients is the implementation of the mouse interface, for manipulating the GUI. With a successfully implemented mouse interface, users can generate mouse events exactly at the point of their choosing. However, it is difficult to define this interaction in the eye-tracking interface. This problem has been defined as the Midas touch problem and has been a major focus of eye-tracking research. There have been many attempts to solve this problem using blink, voice input, etc. However, it was not suitable for general paralyzed patients because some of them cannot wink or speak. In this paper, we propose a mouth-pop-up, eye-tracking mouse interface that solves the Midas touch problem as well as becoming a suitable interface for general paralyzed patients using a common RGB camera. The interface presented in this paper implements a mouse interface that detects the opening and closing of the mouth to activate a pop-up menu that the user can select the mouse event. After implementation, a performance experiment was conducted. As a result, we found that the number of malfunctions and the time to perform tasks were reduced compared to the existing method.