• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.71-73
• /
• 2019

How much and where vehicles exist is an essential element in the implementation of a GeoAI-based urban environment that reflects traffic information. In this paper, we trained vehicle data using Mask R-CNN that deep learning model useful for object detection and extraction, and verified vehicle detection in actual aerial images taken with drones.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.11 no.4
• /
• pp.280-286
• /
• 1986

A newmethod of fast mask operators for edge detection is proposed, which is based on the matrix factorization. The output of each component in the multi-directional mask operator is obtained adding every image pixels in the mask area weighting by corresponding mask element. Therefore, it is same as the result of matrix-vector multiplication like one dimensional transform, i, e, , trasnform of an image vector surrounded by mask with a transform matrix consisted of all the elements of eack mask row by row. In this paper, for the Sobel and Prewitt operators, we find the transform matrices, add up the number of operations factoring these matrices and compare the performances of the proposed method and the standard method. As a result, the number of operations with the proposed method, for Sobel and prewitt operators, without any extra storage element, are reduced by 42.85% and 50% of the standard operations, respectively and in case of an image having 100x100 pixels, the proposed Sobel operator with 301 extra storage locations can be computed by 35.93% of the standard method.

• Journal of Information Processing Systems
• /
• v.18 no.3
• /
• pp.411-427
• /
• 2022

Digging historical and cultural information from seals in ancient books is of great significance. However, ancient Chinese seal samples are scarce and carving methods are diverse, and traditional digital image processing methods based on greyscale have difficulty achieving superior segmentation and recognition performance. Recently, some deep learning algorithms have been proposed to address this problem; however, current neural networks are difficult to train owing to the lack of datasets. To solve the afore-mentioned problems, we proposed an SEL-RefineMask which combines selector of feature pyramid network (SEL-FPN) with RefineMask to segment and recognize seals. We designed an SEL-FPN to intelligently select a specific layer which represents different scales in the FPN and reduces the number of anchor frames. We performed experiments on some instance segmentation networks as the baseline method, and the top-1 segmentation result of 64.93% is 5.73% higher than that of humans. The top-1 result of the SEL-RefineMask network reached 67.96% which surpassed the baseline results. After segmentation, a vision transformer was used to recognize the segmentation output, and the accuracy reached 91%. Furthermore, a dataset of seals in ancient Chinese books (SACB) for segmentation and small seal font (SSF) for recognition were established which are publicly available on the website.

• Journal of Multimedia Information System
• /
• v.8 no.3
• /
• pp.159-166
• /
• 2021

COVID-19 has dramatically changed people's daily life. Wearing masks is considered as a simple but effective way to defend the spread of the epidemic. Hence, a real-time and accurate mask wearing detection system is important. In this paper, a deep learning-based mask wearing detection system is developed to help people defend against the terrible epidemic. The system consists of three important functions, which are image detection, video detection and real-time detection. To keep a high detection rate, a deep learning-based method is adopted to detect masks. Unfortunately, according to the suddenness of the epidemic, the mask wearing dataset is scarce, so a mask wearing dataset is collected in this paper. Besides, to reduce the computational cost and runtime, a simple online and real-time tracking method is adopted to achieve video detection and monitoring. Furthermore, a function is implemented to call the camera to real-time achieve mask wearing detection. The sufficient results have shown that the developed system can perform well in the mask wearing detection task. The precision, recall, mAP and F1 can achieve 86.6%, 96.7%, 96.2% and 91.4%, respectively.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.14 no.4
• /
• pp.323-327
• /
• 2021

This paper produced SUS MASK, which is used for OLEDs, using an automatic liquid management system. The SUS MASK was tested by setting the hole diameter to 0.4 mm. The additive F300 was found to be excellent as the hole diameter was close to 0.4 mm and the error range was measured to be 0.08 on average. And as a result of measuring the weight reduction amount of CuCl2 and FeCl3 according to the change in oxidation-reduction potential (ORP), FeCl3 is relatively sensitive to ORP changes. Experiments were conducted on whether ORP (610 mV) and specific gravity (1.463) were automatically controlled while continuously etching the SUS Mask. Experimental results show that the automatic liquid management system is well controlled because the setting value is not significantly changed. After setting the hole diameter to 0.4 mm as the target, the experiment results were measured from 0.36 to 0.44. Therefore, it is expected that etching processing in the manufacturing process of SUS MASK can be improved with higher precision by applying the manufactured automatic liquid management system.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.7
• /
• pp.113-120
• /
• 2023

The purpose of this study was the accuracy of cardiac compression, fatigue, and physiological changes of the rescuer for different mask type in cardiopulmonary resuscitation(CPR). Data collection was from 9 to 12 May 2022, the participants were a total of 24 paramedic students with a BLS provider at D University. The students participated in an experiment in which 12 students each wore a surgical mask (Dental mask) and a fine particle 94% blocking mask (KF94 mask) and performed CPR for 2 minutes over a total of 7 times. As a result of the study, in the analysis of the quality of the rescuer's chest compression according to the type of mask, there was a significant difference in the compression speed (F=24.91, p<.001) and bad compression hand position (F=14.54, p=.024) in the group wearing the KF94, Fatigue showed significant differences in both the Dental mask group (F=51.16, p<.001) and the KF94 mask group (F=63.49, p<.001). Among the physiological changes, heart rate showed a significant difference between the Dental mask group (F=34.79, p<.001) and the KF94 mask group (F=35.55, p<.001), and the respiratory rate showed a significant difference between the Dental mask group (F=25.02, p=.001) and the KF94 mask group(F=23.02, p=.002). Therefore, in order to improve the quality of efficient chest compression and reduce the fatigue and physiological changes of rescuers, it will be necessary for rescuers to wear suitable personal protective equipment.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.44 no.6
• /
• pp.1163-1177
• /
• 2020

The purpose of this study is to figure out the information needed to improve the shape and size suitability of face-applied mask sheets. The study analyzed the shape of the mask sheet from the scanned images of 50 products of 37 domestic brands. In addition, each measurement of 42 mask sheets were compared and analyzed multilaterally with the 3D measurement dimensions of the faces of men and women in their 20s from the 6th SizeKorea data. Analysis on the shapes of mask sheets indicated that domestic commercial mask sheets are mainly made of single or dual sheets, with slits for enhancing fitness to the three-dimensional face. In the dimensional analysis of Korean men, women and mask sheets, most of the lengths of the mask sheets were significantly larger or smaller than the actual faces of men and women. The horizontal length and vertical length of the forehead above the eyes are significantly shorter, thereby requiring adjustments in the dimensions of this area. In order to improve the size suitability of the mask, it is necessary to adjust the dimensions of the problem area according to the research results as well as diversify the dimensions considering the target layer.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.81-83
• /
• 2024

본 논문에서는 mask R-CNN의 이미지 세그먼테이션(Image Segmentation) 기법을 이용하여 철도의 선로를 식별하고 분류하는 방법을 제안한다. mask R-CNN의 이미지 세그먼테이션은 바운딩 박스(Bounding Box)를 통해 이미지에서 객체를 식별하는 R-CNN 알고리즘과는 달리 픽셀 단위로 관심 있는 객체를 검출하고 분류하는 기법으로서 오브젝트 디텍션(Object Detection)보다 더욱 정교한 객체 식별이 가능하다. 본 연구에서는 Pascal VOC 형태의 고속철도 데이터 24,205셋의 데이터를 전처리하고 MS COCO 데이터셋으로 변환하여, MMDetection의 mask R-CNN을 통해 픽셀 단위로 철도선로를 식별하고 정상/불량 상태를 분류하는 연구를 수행하였다. 선행연구에서는 YOLO를 활용하여 Polygon형태의 좌표를 바운딩 박스로 분류하였는데, 본 연구에서는 mask R-CNN을 활용함으로써 철도 선로를 더욱 정교하게 식별하였으며 정상/불량의 상태 분류는 YOLO와 유사한 성능을 보였다.

• Transactions of the Society of Information Storage Systems
• /
• v.6 no.1
• /
• pp.1-5
• /
• 2010

Holographic data storage system(HDSS) has many noise sources - crosstalk, scattering and inter pixel interference, etc. Generally the intensity of a light generated from the laser source has Gaussian distribution and this ununiformity of light also can make the data page to have a low SNR. A beam apodizer is used to make the laser as a flat-top beam but the intensity distribution is not strictly uniform. The intensity of light can be controlled using image mask. In this paper the intensity distribution of light used for HDSS is controlled by a digital image mask. The digital image mask is changed arbitrarily in real-time with suggested algorithm for the HDSS.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.1
• /
• pp.1-8
• /
• 2016

Convolutional neural network (CNN) consists of a few pairs of both convolution layer and subsampling layer. Thus it has more hidden layers than multi-layer perceptron. With the increased layers, the size of convolution mask ultimately determines the total number of weights in CNN because the mask is shared among input images. It also is an important learning factor which makes or breaks CNN's learning. Therefore, this paper proposes the best method to choose the convolution size and the number of layers for learning CNN successfully. Through our face recognition with vast learning examples, we found that the best size of convolution mask is 5 by 5 and 7 by 7, regardless of the number of layers. In addition, the CNN with two pairs of both convolution and subsampling layer is found to make the best performance as if the multi-layer perceptron having two hidden layers does.