• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.271-274
• /
• 2021

Fisheye 카메라로 촬영된 영상은 일반 영상보다 넓은 시야각을 갖는 장점으로 여러 분야에서 활용되고 있다. 그러나 fisheye 카메라로 촬영된 영상은 어안렌즈의 곡률로 인하여 영상의 중앙 부분은 팽창되고 외곽 부분은 축소되는 방사 왜곡이 발생하기 때문에 영상을 활용함에 있어서 어려움이 있다. 이러한 방사 왜곡을 보정하기 위하여 기존 영상처리 분야에서는 렌즈의 곡률을 수학적으로 계산하여 보정하기도 하지만 이는 각각의 렌즈마다 왜곡 파라미터를 추정해야 하기 때문에, 개별적인 GT (Ground Truth) 영상이 필요하다는 제한 사항이 있다. 이에 본 논문에서는 렌즈의 종류마다 GT 영상을 필요로 하는 기존 기술의 제한 사항을 극복하기 위하여, fisheye 영상만을 입력으로 하여 왜곡계수를 계산하는 딥러닝 네트워크를 제안하고자 한다. 또한, 단일 왜곡계수를 왜곡모델로 활용함으로써 layer 수를 크게 줄일 수 있는 경량화 네트워크를 제안한다.

• Journal of Broadcast Engineering
• /
• v.26 no.5
• /
• pp.608-620
• /
• 2021

Recent deep learning-based face super-resolution (FSR) works showed significant performances by utilizing facial prior knowledge such as facial landmark and dictionary that reflects structural or semantic characteristics of the human face. However, most of these methods require additional processing time and memory. To solve this issue, this paper propose an efficient FSR models using knowledge distillation techniques. The intermediate features of teacher network which contains dictionary information based on major face regions are transferred to the student through adversarial multi-scale features distillation. Experimental results show that the proposed model is superior to other SR methods, and its effectiveness compare to teacher model.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.14 no.4
• /
• pp.35-45
• /
• 2018

Users are sharing many of contents such as text, image, video, and so on in SNS. There are various information as like as personal interesting, opinion, and relationship in social media contents. Therefore, many of recommendation systems or search systems are being developed through analysis of social media contents. In order to extract subject-related topics of social context being collected from social media channels in developing those system, it is necessary to develop ontologies for semantic analysis. However, it is difficult to develop formal ontology because social media contents have the characteristics of non-formal data. Therefore, we develop a social topic system based on semantic and subject correlation. First of all, an extracting system of social topic based on semantic relationship analyzes semantic correlation and then extracts topics expressing semantic information of corresponding social context. Because the possibility of developing formal ontology expressing fully semantic information of various areas is limited, we develop a self-extensible architecture of ontology for semantic correlation. And then, a classifier of social contents and feed back classifies equivalent subject's social contents and feedbacks for extracting social topics according semantic correlation. The result of analyzing social contents and feedbacks extracts subject keyword, and index by measuring the degree of association based on social topic's semantic correlation. Deep Learning is applied into the process of indexing for improving accuracy and performance of mapping analysis of subject's extracting and semantic correlation. We expect that proposed system provides customized contents for users as well as optimized searching results because of analyzing semantic and subject correlation.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.01a
• /
• pp.327-329
• /
• 2022

본 논문에서는 AI Hub의 개방 데이터인 '유형별 두피 이미지'를 사용하여 두피 상태에 대한 신경망을 학습한다. 이 두피 상태에는 6가지 상태가 있는데, 각각의 상태들에 대한 평가를 양호(0)부터 심각(3)까지 분류하여 학습한 신경망 모델로 실제 어플리케이션으로 구현하여 사람들의 두피 사진을 찍어서 두피 상태를 진단한다. 이 과정에서 기존 개방 데이터에서 사용했던 값 비싼 두피 진단기를 사용하는 것이 아닌 값싸게 구할 수 있는 스마트폰용 현미경을 사용하여 좀 더 효율적으로 두피 상태를 진단 할 수 있는 어플리케이션을 만들었다. 몇백만 원 상당의 비싼 두피 진단기로 촬영한 사진과 비교하였을 시 평균적으로 65%의 정확도를 보여주고 있으며 데이터가 많은 유형은 77%의 정확도까지도 보여주었다.

• Journal of Digital Convergence
• /
• v.13 no.9
• /
• pp.259-267
• /
• 2015

In this paper, we proposed the straight line determination method and the algorithm for 8 directions determination of straight line using the coordinate information and the property of trigonometric function. We conduct an experiment that is 8 hand gestures are carried out 100 times each, a total of 800 times. And the accuracy for the 8 derection determination algorithm is showed the diagonal direction to the left upper side shows the highest accuracy as 92%, and the direction to the left side, the diagonal direction to the right upper side and the diagonal direction to the right bottom side show the lowest accuracy as 82%. This method with coordinate information through image processing than the existing recognizer and the recognition through learning process is possible using a hand gesture recognition gesture.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.5
• /
• pp.1001-1010
• /
• 2021

Normal people are not deeply aware of their dependence on sight when eating. However, since the visually impaired do not know what kind of food is on the table, the assistant next to them holds the blind spoon and explains the position of the food in a clockwise direction, front and rear, left and right, etc. In this paper, we describe the development of a meal assistance system that recognizes each food image and announces the name of the food by voice when a visually impaired person looks at their table using a smartphone camera. This system extracts the food on which the spoon is placed through the YOLO model that has learned the image of food and tableware (spoon), recognizes what the food is, and notifies it by voice. Through this system, it is expected that the visually impaired will be able to eat without the help of a meal assistant, thereby increasing their self-reliance and satisfaction.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.1
• /
• pp.33-38
• /
• 2020

This paper proposes a technique for counting construction materials by analyzing an image acquired by a Drone. The proposed technique use drone log which includes drone and camera information, RCNN for predicting construction material type, dummy area and Photogrammetry for counting the number of construction material. The existing research has large error ranges for predicting construction material detection and material dummy area, because of a lack of training data. To reduce the error ranges and improve prediction stability, this paper increases the training data with a method of data augmentation, but only uses rotated training data for data augmentation to prevent overfitting of the training model. For the quantity calculation, we use a drone log containing drones and camera information such as Yaw and FOV, RCNN model to find the pile of building materials in the image and to predict the type. And we synthesize all the information and apply it to the formula suggested in the paper to calculate the actual quantity of material pile. The superiority of the proposed method is demonstrated through experiments.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.5
• /
• pp.619-627
• /
• 2021

Recent advance in low cost and light-weight drones has extended their application areas in both military and private sectors. Accordingly surveillance program against unfriendly drones has become an important issue. Drone detection and classification technique has long been emphasized in order to prevent attacks or accidents by commercial drones in urban areas. Most commercial drones have small sizes and low reflection and hence typical sensors that use acoustic, infrared, or radar signals exhibit limited performances. Recently, artificial intelligence algorithm has been actively exploited to enhance radar image identification performance. In this paper, we adopt machined learning algorithm for high resolution radar imaging in drone detection and classification applications. For this purpose, simulation is carried out against commercial drone models and compared with experimental data obtained through high resolution radar field test.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.12
• /
• pp.57-66
• /
• 2023

In this study, we propose a novel research framework for the recommendation system that can estimate the user's emotional state and reflect it in the recommendation process by applying deep learning techniques and emotion AI (artificial intelligence). To this end, we build an emotion classification model that classifies each of the seven emotions of angry, disgust, fear, happy, sad, surprise, and neutral, respectively, and propose a model that can reflect this result in the recommendation process. However, in the general emotion classification data, the difference in distribution ratio between each label is large, so it may be difficult to expect generalized classification results. In this study, since the number of emotion data such as disgust in emotion image data is often insufficient, correction is made through augmentation. Lastly, we propose a method to reflect the emotion prediction model based on data through image augmentation in the recommendation systems.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.45 no.2
• /
• pp.48-55
• /
• 2022

The color image of the brand comes first and is an important visual element that leads consumers to the consumption of the product. To express more effectively what the brand wants to convey through design, the printing market is striving to print accurate colors that match the intention. In 'offset printing' mainly used in printing, colors are often printed in CMYK (Cyan, Magenta, Yellow, Key) colors. However, it is possible to print more accurate colors by making ink of the desired color instead of dotting CMYK colors. The resulting ink is called 'spot color' ink. Spot color ink is manufactured by repeating the process of mixing the existing inks. In this repetition of trial and error, the manufacturing cost of ink increases, resulting in economic loss, and environmental pollution is caused by wasted inks. In this study, a deep learning algorithm to predict printed spot colors was designed to solve this problem. The algorithm uses a single DNN (Deep Neural Network) model to predict printed spot colors based on the information of the paper and the proportions of inks to mix. More than 8,000 spot color ink data were used for learning, and all color was quantified by dividing the visible light wavelength range into 31 sections and the reflectance for each section. The proposed algorithm predicted more than 80% of spot color inks as very similar colors. The average value of the calculated difference between the actual color and the predicted color through 'Delta E' provided by CIE is 5.29. It is known that when Delta E is less than 10, it is difficult to distinguish the difference in printed color with the naked eye. The algorithm of this study has a more accurate prediction ability than previous studies, and it can be added flexibly even when new inks are added. This can be usefully used in real industrial sites, and it will reduce the attempts of the operator by checking the color of ink in a virtual environment. This will reduce the manufacturing cost of spot color inks and lead to improved working conditions for workers. In addition, it is expected to contribute to solving the environmental pollution problem by reducing unnecessarily wasted ink.