• 한국멀티미디어학회논문지
• /
• 제24권7호
• /
• pp.880-889
• /
• 2021

In everyday life, recognizing people's emotions from their frames is essential and is a popular research domain in the area of computer vision. Visual emotion has a severe class imbalance in which most of the data are distributed in specific categories. The existing methods do not consider class imbalance and used accuracy as the performance metric, which is not suitable for evaluating the performance of the imbalanced dataset. Therefore, we proposed a method for recognizing visual emotion using balanced data augmentation to address the class imbalance. The proposed method generates a balanced dataset by adopting the random over-sampling and image transformation methods. Also, the proposed method uses the Focal loss as a loss function, which can mitigate the class imbalance by down weighting the well-classified samples. EfficientNet, which is the state-of-the-art method for image classification is used to recognize visual emotion. We compare the performance of the proposed method with that of conventional methods by using a public dataset. The experimental results show that the proposed method increases the F1 score by 40% compared with the method without data augmentation, mitigating class imbalance without loss of classification accuracy.

• 소성∙가공
• /
• 제31권4호
• /
• pp.229-239
• /
• 2022

Convolution Neural Network(CNN) is a class of deep learning algorithms and can be used for image analysis. In particular, it has excellent performance in finding the pattern of images. Therefore, CNN is commonly applied for recognizing, learning and classifying images. In this study, the surface defect classification performance of Al 6061 extruded material using CNN-based algorithms were compared and evaluated. First, the data collection criteria were suggested and a total of 2,024 datasets were prepared. And they were randomly classified into 1,417 learning data and 607 evaluation data. After that, the size and quality of the training data set were improved using data augmentation techniques to increase the performance of deep learning. The CNN-based algorithms used in this study were VGGNet-16, VGGNet-19, ResNet-50 and DenseNet-121. The evaluation of the defect classification performance was made by comparing the accuracy, loss, and learning speed using verification data. The DenseNet-121 algorithm showed better performance than other algorithms with an accuracy of 99.13% and a loss value of 0.037. This was due to the structural characteristics of the DenseNet model, and the information loss was reduced by acquiring information from all previous layers for image identification in this algorithm. Based on the above results, the possibility of machine vision application of CNN-based model for the surface defect classification of Al extruded materials was also discussed.

• 한국동물위생학회지
• /
• 제45권3호
• /
• pp.221-228
• /
• 2022

Cataracts occur commonly in older dogs, which can lead to partial or complete vision loss. In the present study, a 12-year-old male poodle presented for evaluation of ocular sinister (OS) cataract. This study was a clinical case report on the process of performing surgery using phacoemulsification (PHACO) and the problems that arise in patient diagnosed with OS cataract. In the oculus dexter (OD), the artificial eye was inserted because there was no electroretinography (ERG) response. In the OS, the ERG was 51.6 μV, so operation was performed because the visual pathway function remained. Slit-lamp biomicroscopy of the OS showed complete loss of vision as a hypermature cataract and that the lens was hardened. During the cataract surgery using PHACO, visco-elastic agents were used to maintain the shape of the eyeball, and the PHACO procedure took 3.13 minutes. The hardened lens and visco-elastic agents were removed from the eye through PHACO surgery, and the operation was completed by inserting an intraocular lens (IOL). As a result of managing IOP for 2 months after surgery, it remained stable at a maximum of 19 mmHg, and no postoperative ocular hypertension (POH) occurred. Currently, one year has passed since the operation, and the dog maintains its daily life with its left eye without any problems.

• 한국환경과학회:학술대회논문집
• /
• 한국환경과학회 2002년도 The 10th Anniversary of KENSS International Workshop
• /
• pp.29-44
• /
• 2002

Wetlands provide vital habitats for fish and wildlife while offering numerous other benefits. As in some other countries, however, Taiwan has witnessed the loss of a significant portion of its coastal wetlands due to large-scale reclamation projects along the coast. Most of the wetlands that still remain are seriously being threatened by both human activities and natural changes, such as drainage for agricultural production, filling for industrial development, discharge of wastes and drought. The administrative performance of the existing authorities and legislation in Taiwan has mostly been ineffective in protecting these precious, sensitive areas. This paper introduces the distribution of wetlands in Taiwan and highlights their invaluable functions and potential economic value. It also discussed the recent activities, both initiated by the government and the NGOs, to protect wetlands in Taiwan. Based on the above discussion, the paper identifies the wide-range of current problems related to their management and proposes the vision should have to save wetlands for the future. It argues that establishing clear policies and effective institutional mechanisms on wetland protection and conservation, classifying the wetlands for better management, and fully promoting public awareness and consciousness of the importance of the wetlands will not only be beneficial but will also address the urgent need to safeguard the wetlands in Taiwan. Additionally, the paper recognizes that international cooperation and collaboration on wetland restoration is essential and most challenging.

• 국제학술발표논문집
• /
• The 6th International Conference on Construction Engineering and Project Management
• /
• pp.313-316
• /
• 2015

Construction is one of the most dangerous job sectors, which reports tens of thousands of time-loss injuries and deaths every year. These disasters incur delays and additional costs to the projects. The safety management needs to be on the top primary tasks throughout the construction to avoid fatal accidents and to foster safe working environments. One of the safety regulations that are frequently violated is the wearing of personal protection equipment (PPE). In order to facilitate monitoring of the compliance of the PPE wearing regulations, this paper proposes a vision based method that automatically identifies whether workers wear hard hats and safety vests. The method involves three modules - human body detection, identification of safety vest wearing, and hard hat detection. First, human bodies are detected in the video frames captured by real-time on-site construction cameras. The detected human bodies are classified into with/without wearing safety vests based on the color features of their upper parts. Finally, hard hats are detected on the nearby regions of the detected human bodies and the locations of the detected hard hats and human bodies are correlated to reveal their corresponding matches. In this way, the proposed method provides any appearance of the workers without wearing hard hats or safety vests. The method has been tested on onsite videos and the results signify its potential to facilitate site safety monitoring.

• 한국실내디자인학회논문집
• /
• 제23권2호
• /
• pp.71-79
• /
• 2014

The purpose of this study is to elucidate that photo, which has been used as original photo's purpose of information transfer in modern age, is now used as romantic reproduction which is the communication methods of architect's idea and thought through photographic vision which is beyond photograph own capabilities. The photos of Mies van der Rohe and Le Corbusier's architectural works are taken as an example for studying and analysing the way of deliverying the concept of creative work in the functional spaces in the modern era. It looked at the way of modern archirecture configuration, which architects wanted to show by pictures, such as concurrency, movement, sense of exhibition and concept of time-space and planarity on photographic vision such as multiview, movement, daily life exclusion, scenography and perspective loss. Reflecting that, I presents Le Corbusier and Mies van der Rohe's intention through photo by analyzing their picture of architecture by way of photograph techniques-camera position moving, over exposure, photomontage, silhouette technic and overlap technic. Mies van der Rohe and Le corbusier demonstrated the change and manipulation of the their architectural photos in different point of view. They express their architectural theories by photos of their works and overcome the limitation of expression of constructed building designed by them. The photos of architects's works in the case study with photos and descriptions introduce to their design concept. The design concept of the architects have become ideal concept for many contemporary architects and keep reproducing through the photos of their architectural works.

• 한국임상수의학회지
• /
• 제25권5호
• /
• pp.385-390
• /
• 2008

The purpose of this study was to compare the long term prognosis of Ahmed valve implantation and medical therapy in dogs with early primary glaucoma. The medical records of dogs with early primary glaucoma from the Veterinary Medical Teaching Hospital of Seoul National University between January 2005 and May 2008 were reviewed. Five dogs underwent Ahmed valve implantation and medical therapy was performed in six dogs. Breeds of the dogs included Shih Tzu(n = 7), American cocker spaniel(n = 2), and miniature poodle(n = 2). Adequate intraocular pressure(IOP) was better maintained in the Ahmed valve implantation group than in the medical treatment group throughout the follow-up period. Mean time for vision loss to develop after treatment in the Ahmed valve implantation group($198.0{\pm}115.4$ days) was significantly longer than that in medical treatment group($43.2{\pm}30.4$ days, p < 0.05). Post-operative complications included occlusion of the tube by fibrin in the anterior chamber (n = 1), fibrous capsule about the episcleral base of the device(n = 1), and dislocation of the valve body(n = 1). It is considered that Ahmed valve implantation is more effective strategy to control IOP and maintain vision than medical treatment in early glaucomatous dogs.

• 한국통신학회논문지
• /
• 제38A권9호
• /
• pp.773-781
• /
• 2013

허프 변환은 데이터 손실 및 왜곡이 포함된 영상에서도 직선 정보 추출에 용이한 장점이 있어 컴퓨터 비전 분야의 응용분야에 널리 사용되어 왔다. 그러나 허프 변환의 보팅 과정은 비효율적인 연산구조와 많은 메모리 접근횟수로 인해 실시간 처리 임베디드 비전 시스템에 적용하는데 한계가 있다. 이에 본 논문에서는 허프 변환의 개선된 보팅 정책을 제시하고, 이를 적용하여 적은 하드웨어 자원 사용량으로 실시간 성능을 만족하는 허프 변환의 하드웨어 구조를 제안한다. 제안된 보팅 정책은 인접한 픽셀 간의 관계를 이용하여 보팅 연산 과정의 오버헤드를 줄였으며, 하드웨어 재사용성을 높임으로서 효율적인 연산구조를 가진다. 이러한 개선된 보팅 정책을 적용한 제안된 하드웨어 구조는 인접한 픽셀들의 보트 값을 병렬적으로 연산하고 저장하여 시간당 처리량을 높인다. 제안 구조의 장점은 순차적 연산 방식 대비 매우 적은 추가 하드웨어 자원만으로 이러한 성능 향상을 위한 병렬화를 달성한다는 것이다.

• 한국산학기술학회논문지
• /
• 제9권3호
• /
• pp.605-610
• /
• 2008

용접구조물의 시공시 발견되는 외관상, 성능상 사용자에게 불만족을 발생시키는 결함들을 일반적으로 총칭하여 용접부의 결함이라 한다. 이들 결함들을 시간의 손실 없이 효과적으로 검출해내어 용접부 전체의 품질을 효과적으로 판정하기 위한 용접성 평가 시스템 구축이 어느 때보다 절실하다. 본 연구에서는 레이저 비전 카메라를 이용하여, 용접부의 형상에 대한 원 데이터를 채취하고 이를 화상처리하고, 여기서 얻은 정보로 1차적으로 정량적 결함을 판별하고, 동시에 분류된 결함을 정성적으로 판단하기 위해 퍼지 패턴인식의 한 단계인 다중 패턴 인식을 이용하여 용접 물 전체의 결함을 판정하였으며, 사용자의 편의를 위해 용접성의 결과가 화상별, 최종집계와 함께 이를 그래픽의 형태로 표시하여, 손쉽게 용접성 양부를 결정할 수 있게 하였으며, 이는 또한 생산라인에 적용하여 실시간 용접성 판별을 할 수 있는 기반 연구 성과를 제시하였다.

• 동의신경정신과학회지
• /
• 제16권2호
• /
• pp.287-299
• /
• 2005

Dissociative disorder is a psychiatric disorder characterized by a sudden loss of memory, but which has no organic disease or explanation. It usually occurs after heavy psychosocial stress or traumatic experience. A transient cerebral ischemic attack (TIA) is an acute episode of temporary and focal loss of cerebral function of vascular origin. TIAs are rapid in onset; symptoms reach their maximal manifestation in fewer than 5 minutes. Manifestations are of variable duration and typically last 2-15 minutes(rarely as long as 24 h). Most TIA durations are less than 1 hour. Of concern is the careful detection of changes in behavior, speech, gait, memory, movement, and vision. TIAs are uncommon in persons younger than 60 years. I treat 6 cases of Sudden Temporary Amnesia Patients with oriental medicine and they are improved. All of them had amnesia for $6{\sim}10\;hours$. During that time, they show behavioral changes and they are not on the state of unconsciousness. After recovery, they also forget what happen at the time. they have some emotional reason too. In conclusion, 4 cases of them belong to dissociative disorder and 2 other cases, TIA.