• Polymer(Korea)
• /
• v.36 no.4
• /
• pp.525-530
• /
• 2012

In this study, silane-crosslinked organic/inorganic composite membranes were prepared by simultaneous irradiation grafting of binary monomer mixtures (styrene and 3-(trimethoxysilyl)propyl methacrylate (TMSPM)) with various compositions onto a poly(ethylene-alt-tetraethylene) (ETFE) film and followed by sol-gel processing and sulfonation to provide a silane-crosslinked structure and a proton conducting ability, respectively. The Fourier transform infrared spectroscopy (FTIR) and thermo gravimetric analysis (TGA) were utilized to confirm the crosslinking of ETFE-g-PS/PTMSPM films. The prepared membranes with similar ion exchange capacity but a different TMSPM content were selected and their membrane properties were compared. The ETFE-g-PSSA/PTMSPM membranes were characterized by water uptake, dimensional stability, and proton conductivity after sulfonation. The membrane electrode assemblies (MEA) of the prepared membranes were fabricated and their single cell performances were measured.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.2
• /
• pp.1414-1419
• /
• 2015

A method for detecting a skin region on the image is generally used to detect the color information. However, If saturation lowered, skin detection is difficult because hue information of the pixels is lost. So in this paper, we propose a method of correcting color of lower saturation of skin region images by the lighting. Color correction process of this method is saturation image acquisition and low-saturation region classification, segmentation, and the saturation of the split in the low saturation region extraction and color values, the color correction sequence. This method extracts the low saturation regions in the image and extract the color and saturation in the region and the surrounding region to produce a color similar to the original color. Therefore, the method of extracting the low saturation region should be correctly preceding. Because more accurate segmentation in the process of obtaining a low saturation regions, we use a multi-threshold method proposed Otsu in Hue values of the HSV color space, and create a binary image. Our experimental results for 170 portrait images show a possibility that the proposed method could be used efficiently preprocessing of skin color detection method, because the detection result of proposed method is 5.8% higher than not used it.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.8
• /
• pp.317-326
• /
• 2013

In this paper, we present an effective method for comparing control flow graphs which represent static structures of binary programs. To compare control flow graphs, we measure similarities by comparing instructions and syntactic information contained in basic blocks. In addition, we also consider similarities of edges, which represent control flows between basic blocks, by edge extension. Based on the comparison results of basic blocks and edges, we match most similar basic blocks in two control flow graphs, and then calculate the similarity between control flow graphs. We evaluate the proposed edge extension method in real world Java programs with respect to structural similarities of their control flow graphs. To compare the performance of the proposed method, we also performed experiments with a previous structural comparison for control flow graphs. From the experimental results, the proposed method is evaluated to have enough distinction ability between control flow graphs which have different structural characteristics. Although the method takes more time than previous method, it is evaluated to be more resilient than previous method in comparing control flow graphs which have similar structural characteristics. Control flow graph can be effectively used in program analysis and understanding, and the proposed method is expected to be applied to various areas, such as code optimization, detection of similar code, and detection of code plagiarism.

• The KIPS Transactions:PartB
• /
• v.13B no.2 s.105
• /
• pp.77-82
• /
• 2006

Digital halftoning is the technique to obtain a bilevel-toned image from continuous-toned image. Among halftoning methods, the error diffusion method gives better subjective quality than other halftoning ones. But it also makes edges of objects blurred. To overcome the defect, we proposes the modified error diffusion to enhance the edges using the property that human vision perceives the local average luminance and doesn't perceive a little variation of the spatial variation. The proposed method computes a spatialactivity, which is the difference between a pixel luminance and the average of its $3{\times}3$ neighborhood pixels' Iuminance weighted according to the spatial positioning. The system also usesof edge enhancement (IEE), which is computed from the normalized spatial activitymultiplied by the average luminance. The IEE is added to the quantizer's input pixel and feeds into the halftoning quantizer. The quantizer produces the halftone image having the enhanced edge. The computer experimental results show that the proposed method produces clearer bilevel-toned images than conventional methodsand the edge of objects is preserved well. Also the performance of the preposed method is improved, compared with that of the conventional method by measuring the edge correlation and the local average accordance at some ranges of viewing distance.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.9
• /
• pp.359-366
• /
• 2021

A number of construction companies in Korea invest considerable human and financial resources to construct a system for managing apartment defect data and for categorizing repair tasks. Thus, this study proposes machine learning models to automatically classify defect complaint text-data into one of the sub categories of 'finishing work' (i.e., one of the defect repair tasks). In the proposed models, we employed two word representation methods (Bag-of-words, Term Frequency-Inverse Document Frequency (TF-IDF)) and two machine learning classifiers (Support Vector Machine, Random Forest). In particular, we conducted both binary- and multi- classification tasks to classify 9 sub categories of finishing work: home appliance installation work, paperwork, painting work, plastering work, interior masonry work, plaster finishing work, indoor furniture installation work, kitchen facility installation work, and tiling work. The machine learning classifiers using the TF-IDF representation method and Random Forest classification achieved more than 90% accuracy, precision, recall, and F1 score. We shed light on the possibility of constructing automated defect classification systems based on the proposed machine learning models.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.11
• /
• pp.41-49
• /
• 2021

All application programs, including malware, call the Application Programming Interface (API) upon execution. Recently, using those characteristics, attempts to detect and classify malware based on API Call information have been actively studied. However, datasets containing API Call information require a large amount of computational cost and processing time. In addition, information that does not significantly affect the classification of malware may affect the classification accuracy of the learning model. Therefore, in this paper, we propose a method of extracting a essential feature set after reducing the dimensionality of API Call information by applying various feature selection methods. We used CICAndMal2020, a recently announced Android malware dataset, for the experiment. After extracting the essential feature set through various feature selection methods, Android malware classification was conducted using CNN (Convolutional Neural Network) and the results were analyzed. The results showed that the selected feature set or weight priority varies according to the feature selection methods. And, in the case of binary classification, malware was classified with 97% accuracy even if the feature set was reduced to 15% of the total size. In the case of multiclass classification, an average accuracy of 83% was achieved while reducing the feature set to 8% of the total size.

• 재활복지
• /
• v.18 no.4
• /
• pp.187-205
• /
• 2014

Is to identify how this affects the physically disabled to participate in the program 12 weeks Park Golf psychological well-being and happiness, the purpose of this research is subjective. How to study subjects, only 40-year-old disabled man more than 24 people total delay experimental group and 12 patients(failure cut seven, delayed dysfunction 5) and the control group and 12 patients(failure cut six, delayed dysfunction in 4, two people were involved in the joint disorder). 3 times a week(Mon, Wed, Fri), was carried out 50 minutes into 12 weeks of the experimental period, was located at River Park Golf Course A test place. We calculate the pre-and post-test data mean and standard deviation using SPSS Statistics 21.0 statistical data processing program, binary repeated measures ANOVA to analyze the effects on the psychological well-being of the disabled and subjective effects euphoria Park Golf Participation(was performed 2-way [2] RM ANOVA). First results in psychological well-being of the two groups according to Park Golf participate in group comparisons before and after the exercise involved only fun, immersive and shows were not significantly different, within each group enjoyment, competence, self-realization, all the children of the immersion showed a significant difference in the factors. Second, before and after participation in exercise, there was a significant difference between groups in subjective happiness of two groups according to Park Golf participation, the two groups were not significantly different within. Taken together the results to see more, showed that the positive effects on the psychological well-being and subjective happiness Park Golf participation is the Physically Disabled.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.11
• /
• pp.435-442
• /
• 2018

Outlier detection means to detect data samples that deviate significantly from the distribution of normal data. Most outlier detection methods calculate an outlier score that indicates the extent to which a data sample is out of normal state and determine it to be an outlier when its outlier score is above a given threshold. However, since the range of an outlier score is different for each data and the outliers exist at a smaller ratio than the normal data, it is very difficult to determine the threshold value for an outlier score. Further, in an actual situation, it is not easy to acquire data including a sufficient amount of outliers available for learning. In this paper, we propose a clustering-based outlier detection method by constructing a model representing a normal data region using only normal data and performing binary classification of outliers and normal data for new data samples. Then, by dividing the given normal data into chunks, and constructing a clustering model for each chunk, we expand it to the ensemble method combining the decision by the models and apply it to the streaming data with dynamic changes. Experimental results using real data and artificial data show high performance of the proposed method.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.16 no.6
• /
• pp.63-71
• /
• 2006

In this paper, we propose a secure solution for user authentication by using fingerprint verification on the sensor-client-server model, even with the client that is not necessarily trusted by the sensor holder or the server. To protect possible attacks launched at the untrusted client, our solution makes the fingerprint sensor validate the result computed by the client for the feature extraction. However, the validation should be simple so that the resource-constrained fingerprint sensor can validate it in real-time. To solve this problem, we separate the feature extraction into binarization and minutiae extraction, and assign the time-consuming binarization to the client. After receiving the result of binarization from the client, the sensor conducts a simple validation to check the result, performs the minutiae extraction with the received binary image from the client, and then sends the extracted minutiae to the server. Based on the experimental results, the proposed solution for fingerprint verification can be performed on the sensor-client-server model securely and in real-time with the aid of an untrusted client.

• 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.