• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.4
• /
• pp.980-997
• /
• 2024

To improve the brightness of images and reveal hidden information in dark areas is the main objective of low-light image enhancement (LLIE). LLIE methods based on deep learning show good performance. However, there are some limitations to these methods, such as the complex network model requires highly configurable environments, and deficient enhancement of edge details leads to blurring of the target content. Single-scale feature extraction results in the insufficient recovery of the hidden content of the enhanced images. This paper proposed an edge detection-based multi-scale feature enhancement network for LLIE (EDMFEN). To reduce the loss of edge details in the enhanced images, an edge extraction module consisting of a Sobel operator is introduced to obtain edge information by computing gradients of images. In addition, a multi-scale feature enhancement module (MSFEM) consisting of multi-scale feature extraction block (MSFEB) and a spatial attention mechanism is proposed to thoroughly recover the hidden content of the enhanced images and obtain richer features. Since the fused features may contain some useless information, the MSFEB is introduced so as to obtain the image features with different perceptual fields. To use the multi-scale features more effectively, a spatial attention mechanism module is used to retain the key features and improve the model performance after fusing multi-scale features. Experimental results on two datasets and five baseline datasets show that EDMFEN has good performance when compared with the stateof-the-art LLIE methods.

• Journal of Broadcast Engineering
• /
• v.25 no.6
• /
• pp.944-953
• /
• 2020

Object detection plays a crucial role in a self-driving system. With the advances of image recognition based on deep convolutional neural networks, researches on object detection have been actively explored. In this paper, we proposed a lightweight model of the mask R-CNN, which has been most widely used for object detection, to efficiently predict location and shape of various objects on the road environment. Furthermore, feature maps are adaptively re-calibrated to improve the detection performance by applying an attention module to the neural network layer that plays different roles within the mask R-CNN. Various experimental results for real driving scenes demonstrate that the proposed method is able to maintain the high detection performance with significantly reduced network parameters.

• Transactions on Electrical and Electronic Materials
• /
• v.18 no.1
• /
• pp.30-34
• /
• 2017

The building integrated photovoltaic system (BIPV) attracts attention with regard to the future of the photovoltaic (PV) industry. It is because one of the promising national and civilian projects in the country. Since land area is limited, there is considerable interest in BIPV systems with a variety of angles and shapes of PV panels. It is therefore expected to be one of the major fields for the PV industry in the future. Since the irradiation is different from each installation angle, the output can be predicted by the angles. This is critical for a PV system to be operated at maximum power and use an efficient design. The development characteristics of tilted angles based on data results obtained via long-term monitoring need to be analyzed. The ratio of the theoretically available and actual outputs is compared with the installation angles of each PV module to provide a suitable PV system for the user.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.295-296
• /
• 2022

작물을 경작 해충과 질병은 오랫동안 주요 관심사였다. 농업에서 병해충을 탐지하기 위해 전통적인 방법을 사용하는 것은 더 이상 높은 효율성을 제공하지 않는다. 오늘날 과학과 인공 지능의 폭발적인 발달로 인해 농업분야의 연구원들은 병해충을 탐지하기 위해 딥 러닝을 적용하고 있다. 최근에 다양한 분야의 문제들을 해결하기 위해 수많은 모델들이 발표되었지만, 많은 병해충 진단 딥러닝을 사용한 방법들은 하드웨어 리소스를 낭비하고 실제 농장에서 사용하기 어렵다. 따라서 본 논문에서는 작물의 병해충을 분류하기 위해 Select Kernel Attention(SK Attention)을 Channel Attention Module 로 변경하여 Decoupling-and-Attention network (DANet)을 하드웨어 리소스 사용을 최소화한다.

• Smart Structures and Systems
• /
• v.29 no.1
• /
• pp.1-16
• /
• 2022

Despite recent breakthroughs in deep learning and computer vision fields, the pixel-wise identification of tiny objects in high-resolution images with complex disturbances remains challenging. This study proposes a modified U-net for tiny crack segmentation in real-world steel-box-girder bridges. The modified U-net adopts the common U-net framework and a novel Self-Attention-Self-Adaption (SASA) neuron as the fundamental computing element. The Self-Attention module applies softmax and gate operations to obtain the attention vector. It enables the neuron to focus on the most significant receptive fields when processing large-scale feature maps. The Self-Adaption module consists of a multiplayer perceptron subnet and achieves deeper feature extraction inside a single neuron. For data augmentation, a grid-based crack random elastic deformation (CRED) algorithm is designed to enrich the diversities and irregular shapes of distributed cracks. Grid-based uniform control nodes are first set on both input images and binary labels, random offsets are then employed on these control nodes, and bilinear interpolation is performed for the rest pixels. The proposed SASA neuron and CRED algorithm are simultaneously deployed to train the modified U-net. 200 raw images with a high resolution of 4928 × 3264 are collected, 160 for training and the rest 40 for the test. 512 × 512 patches are generated from the original images by a sliding window with an overlap of 256 as inputs. Results show that the average IoU between the recognized and ground-truth cracks reaches 0.409, which is 29.8% higher than the regular U-net. A five-fold cross-validation study is performed to verify that the proposed method is robust to different training and test images. Ablation experiments further demonstrate the effectiveness of the proposed SASA neuron and CRED algorithm. Promotions of the average IoU individually utilizing the SASA and CRED module add up to the final promotion of the full model, indicating that the SASA and CRED modules contribute to the different stages of model and data in the training process.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.24 no.2
• /
• pp.175-186
• /
• 2021

In the field of automatic target recognition(ATR) with synthetic aperture radar(SAR) imagery, it is usually impractical to obtain SAR target images covering a full range of aspect views. When the database consists of SAR target images with limited angular diversity, it can lead to performance degradation of the SAR-ATR system. To address this problem, this paper proposes a deep learning-based method where channel attention modules(CAMs) are inserted to a convolutional neural network(CNN). Motivated by the idea of the squeeze-and-excitation(SE) network, the CAM is considered to help improve recognition performance by selectively emphasizing discriminative features and suppressing ones with less information. After testing various CAM types included in the ResNet18-type base network, the SE CAM and its modified forms are applied to SAR target recognition using MSTAR dataset with different reduction ratios in order to validate recognition performance improvement under the limited angular diversity condition.

• Journal of The Korean Association of Information Education
• /
• v.21 no.2
• /
• pp.191-198
• /
• 2017

The purpose of this study is to design and apply cooperative learning using module cards for SW education for beginner and to analyze their effects and perceptions. Cooperative learning using 30 module cards consists of 3 activities and we applied it to elementary school students to verify the effectiveness of the learning. For the research analysis, we analyzed the pre-post motivation of SW education, the degree of satisfaction, interest level, and step recognition of the cooperative learning. As a result of the analysis, the learning motivation showed significant improvement in all areas of attention, relevance, confidence, and satisfaction. Students also found that the second stage of cooperative learning using module cards was the most interesting and the third was the most helpful.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.2
• /
• pp.349-355
• /
• 2020

Social robots are attracting attention as an alternative to social problems in modern society, and the need for modularization has been raised to efficiently manage various robots. The study aims to reestablish the concept of 'Personal social service robot' and propose a new modularization method to apply it. For this study, literature research about the definition of social robots and service robots is conducted. In addition, we investigated the modularization of robots and analyzed the modularization cases of service robots. Based on this, we have deducted considerations for applying social service robot modularization and proposed a new modularization. This study divided the module into active module and passive module according to whether it is electric or electronic component of the module, and the active module was again classified into basic module and additional module according to the basic and replacement type of the robot. The modularization was verified by making the prototype of the actual robot.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.67 no.3
• /
• pp.137-142
• /
• 2018

Energy harvesting technology is drawing attention as a means of collecting various eco-friendly energy and accumulating residual energy. Recently, differential power processing (DPP) is being developed as part of energy harvesting. This is being studied as a solution to the loss of power generation between power modules and the problems caused by module small losses depending on the size of power production. In this paper, we propose the necessity of the DPP by comparing and analyzing energy harvesting related module integration system and power supply efficiency of DPP. The power efficiency of the converter and the power difference between the wind power and the photovoltaic power supply have been changed to demonstrate the effectiveness of the proposed system.

• International Journal of Internet, Broadcasting and Communication
• /
• v.13 no.1
• /
• pp.90-99
• /
• 2021

Recently, studies on wearable devices in ubiquitous computing environments have increased and the technology collecting user's activities to provide services has received great attention. We have compared the step counting methods according to sensor molding methods in case of counting steps by using the piezoelectric sensor. We have classified the cases which could result from the course of molding the internal module of a smart shoe as follows: (i) the module is unmolded, (ii) molded but only to the extent that a sensor is fixed or (iii) molded to the extent that a sensor is not moved. Moreover, we have made comparison to verify which algorithm should be used to increase the accuracy of counting steps by the respective cases. Based on the comparison result, we have confirmed that the accuracy of counting steps is higher when using gradient value rather than when using threshold value. In the case of no molding and small molding under the condition of using gradient value, it was turned out to be 100% accuracy for step counting.