• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.19 no.2
• /
• pp.435-442
• /
• 2015

In this paper, we propose a tracking and recognition of pedestrian/vehicle for intelligent multi-visual surveillance system. The intelligent multi-visual surveillance system consists of several fixed cameras and one calibrated PTZ camera, which automatically tracks and recognizes the detected moving objects. The fixed wide-angle cameras are used to monitor large open areas, but the moving objects on the images are too small to view in detail. But, the PTZ camera is capable of increasing the monitoring area and enhancing the image quality by tracking and zooming in on a target. The proposed system is able to determine whether the detected moving objects are pedestrian/vehicle or not using the SVM. In order to reduce the tracking error, an improved camera calibration algorithm between the fixed cameras and the PTZ camera is proposed. Various experimental results show the effectiveness of the proposed system.

• Journal of Information Processing Systems
• /
• v.14 no.1
• /
• pp.162-175
• /
• 2018

Currently, electricity consumption and feedback mechanisms are being widely researched in Internet of Things (IoT) areas to realise power consumption monitoring and management through the remote control of appliances. This paper aims to develop a smart electricity utilisation IoT platform with a deep belief network for electricity utilisation feature modelling. In the end node of electricity utilisation, a smart monitoring and control module is developed for automatically operating air conditioners with a gateway, which connects and controls the appliances through an embedded ZigBee solution. To collect electricity consumption data, a programmable smart IoT gateway is developed to connect an IoT cloud server of smart electricity utilisation via the Internet and report the operational parameters and working states. The cloud platform manages the behaviour planning functions of the energy-saving strategies based on the power consumption features analysed by a deep belief network algorithm, which enables the automatic classification of the electricity utilisation situation. Besides increasing the user's comfort and improving the user's experience, the established feature models provide reliable information and effective control suggestions for power reduction by refining the air conditioner operation habits of each house. In addition, several data visualisation technologies are utilised to present the power consumption datasets intuitively.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.11a
• /
• pp.798-801
• /
• 2014

여행 정보 시스템(ATIS), 교통 관리 시스템 (ITS) 등 궤적 기반 서비스에서, 서비스 품질을 향상시키기 위해서는 주어진 궤적 질의에 대한 정확한 주행시간을 예측하는 것이 필수적이다. 이를 위한 대표적인 공간 데이터 분석 기법으로는 데이터 분류에서 높은 정확도를 보장하는 규칙 기반 분류화 기법이 존재한다. 그러나 기존 규칙 기반 분류화 기법은 단일 컴퓨터 환경만을 고려하기 때문에, 대용량 공간 데이터 처리에 적합하지 않은 문제점이 존재한다. 이를 해결하기 위해, 본 연구에서는 맵리듀스 환경에서 규칙 기반 분류화를 이용한 궤적 데이터 주행 시간 예측 알고리즘을 개발하고자 한다. 제안하는 알고리즘은 첫째, 맵리듀스를 이용하여 대용량 공간 데이터를 병렬적으로 분석함으로써, 활용도 높은 궤적 데이터 규칙을 생성한다. 이를 통해 대용량 공간 데이터 기반의 규칙 생성 시간을 감소시킨다. 둘째, 그리드 구조 기반의 지도 데이터 분할을 통해, 사용자 질의처리 시 탐색 성능을 향상시킨다. 즉, 주행 시간 예측을 위한 규칙 그룹을 탐색 시 질의를 포함하는 그리드 셀만을 탐색하기 때문에, 질의처리 성능이 향상된다. 마지막으로 맵리듀스 구조에 적합한 질의처리 알고리즘을 설계하여, 효율적인 병렬 질의처리를 지원한다. 이를 위해 맵 함수에서는 선정된 그리드 셀에 대해, 질의에 포함된 도로 구간에서의 주행 시간을 병렬적으로 측정한다. 아울러 리듀스 함수에서는 출발 시간 및 구간별 주행 시간을 바탕으로 맵 함수의 결과를 병합함으로써, 최종 결과를 생성한다. 이를 통해 공간 빅데이터 분석을 통한 주행 시간 예측 기법의 처리 시간 및 결과 정확도를 향상시킨다.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.02a
• /
• pp.413-413
• /
• 2010

In semiconductor manufacturing field, all equipments have various sensors to diagnosis the situations of processes. For increasing the accuracy of diagnosis, hundreds of sensors are emplyed. As sensors provide millions of data, the process diagnosis from them are unrealistic. Besides, in some cases, the results from some data which have same conditions are different. We want to find some information, such as data and knowledge, from the data. Nowadays, fault detection and classification (FDC) has been concerned to increasing the yield. Certain faults and no-faults can be classified by various FDC tools. The uncertainty in semiconductor manufacturing, no-faulty in faulty and faulty in no-faulty, has been caused the productivity to decreased. From the uncertainty, the rough set theory is a viable approach for extraction of meaningful knowledge and making predictions. Reduction of data sets, finding hidden data patterns, and generation of decision rules contrasts other approaches such as regression analysis and neural networks. In this research, a RGB sensor was used for diagnosis plasma instead of optical emission spectroscopy (OES). RGB data has just three variables (red, green and blue), while OES data has thousands of variables. RGB data, however, is difficult to analyze by human's eyes. Same outputs in a variable show different outcomes. In other words, RGB data includes the uncertainty. In this research, by rough set theory, decision rules were generated. In decision rules, we could find the hidden data patterns from the uncertainty. RGB sensor can diagnosis the change of plasma condition as over 90% accuracy by the rough set theory. Although we only present a preliminary research result, in this paper, we will continuously develop uncertainty problem solving data mining algorithm for the application of semiconductor process diagnosis.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.10
• /
• pp.1010-1014
• /
• 2010

Article-related product recommender system is an emerging e-commerce service which recommends items based on association in contexts between items and articles. Current services recommend based on the similarity between tags of articles and items, which is deficient not only due to the high cost in manual tagging but also low accuracies in recommendation. As a component of novel article-related item recommender system, we propose a new method for tagging item images based on pre-defined categories. We suggest a hypernetwork-based algorithm for learning association between images, which is represented by visual words, and categories of products. Learned hypernetwork are used to assign multiple tags to unlabeled item images. We show the ability of our method with a product set of real-world online shopping-mall including 1,251 product images with 10 categories. Experimental results not only show that the proposed method has competitive tagging performance compared with other classifiers but also present that the proposed multi-tagging method based on hypernetworks improves the accuracy of tagging.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.7
• /
• pp.1387-1393
• /
• 2007

Even if the linearly separable patterns can be classified by the conventional single layer perceptron, the non-linear problems such as XOR can not be classified by it. A fuzzy single layer perceptron can solve the conventional XOR problems by applying fuzzy membership functions. However, in the fuzzy single layer perception, there are a couple disadvantages which are a decision boundary is sometimes vibrating and a convergence may be extremely lowered according to the scopes of the initial values and learning rates. In this paper, for these reasons, we proposed an enhanced fuzzy single layer perceptron algorithm that can prevent from vibration the decision boundary by introducing a bias term and can also reduce the learn time by applying the modified delta rule which include the learning rates and the momentum concept and applying the new linear activation function. Consequently, the simulation results of the XOR and pattern classification problems presented that the proposed method provided the shorter learning time and better convergence than the conventional fuzzy single layer perceptron.

• Korean Journal of Remote Sensing
• /
• v.24 no.2
• /
• pp.117-124
• /
• 2008

This paper introduces a novel approach for automated mapping of a map feature that is vinyl green house in high spatial resolution imagery Some map features have their unique spatial patterns. These patterns are normally detected in high spatial resolution remotely sensed data by human recognition system. When spatial patterns can be applied to map feature identification, it will improve image classification accuracy and will be contributed a lot to feature identification. In this study, an automated feature identification approach using spatial aucorrelation is developed, specifically for the vinyl green house that has distinctive spatial pattern in its array. The algorithm aimed to develop the method without any human intervention such as digitizing. The method can investigate the characteristics of repeated spatial pattern of vinyl green house. The repeated spatial pattern comes from the orderly array of vinyl green house. For this, object-based approaches are essential because the pattern is recognized when the shapes that are consists of the groups of pixels are involved. The experimental result shows very effective vinyl house extraction. The targeted three vinyl green houses were exactly identified in the IKONOS image for a part of Jeju area.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.152-152
• /
• 2019

This study demonstrates a novel approach of remotely sensed estimates of stream flow at fifteen hydrological station in the Han River Basin, Korea. Multi-temporal data of the European Space Agency's Sentinel-1 SAR satellite from 19 January, 2015 to 25 August, 2018 is used to develop and validate the flow estimation model for each station. The flow estimation model is based on a power law relationship established between the remotely sensed surface area of water at a selected reach of the stream and the observed discharge. The satellite images were pre-processed for thermal noise, radiometric, speckle and terrain correction. The difference in SAR image brightness caused by the differences in SAR satellite look angle and atmospheric condition are corrected using the histogram matching technique. Selective area filtering is applied to identify the extent of the selected stream reach where the change in water surface area is highly sensitive to the change in stream discharge. Following this, an iterative procedure called the Optimum Threshold Classification Algorithm (OTC) is applied to the multi-temporal selective areas to extract a series of water surface areas. It is observed that the extracted water surface area and the stream discharge are related by the power law equation. A strong correlation coefficient ranging from 0.68 to 0.98 (mean=0.89) was observed for thirteen hydrological stations, while at two stations the relationship was highly affected by the hydraulic structures such as dam. It is further identified that the availability of remotely sensed data for a range of discharge conditions and the geometric properties of the selected stream reach such as the stream width and side slope influence the accuracy of the flow estimation model.

• Journal of Korea Water Resources Association
• /
• v.54 no.10
• /
• pp.759-768
• /
• 2021

This work introduces a new approach that classifies individual household water usage by examining the characteristics of smart meter end-user demand data. Here, one of the most well-known unsupervised machine learning, K-means algorithm, is applied to classify water consumptions by each household. The intensity and duration of end-user demands are used as main features to determine the households with similar water consumption pattern. The results showed that 21 households are classified into 13 clusters with each cluster having one, two, three, or five houses. The reasoning why multiple households are classified into the same cluster is described in this paper with respect to the collected data and end-user water consumption behavior.

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.167-174
• /
• 2021

Recently, deep learning technology is widely used in various computer vision applications, such as object recognition, classification, and image generation. In particular, the deep learning-based super-resolution has been gaining significant performance improvement. Fast super-resolution convolutional neural network (FSRCNN) is a well-known model as a deep learning-based super-resolution algorithm that output image is generated by a deconvolutional layer. In this paper, we propose an FPGA-based convolutional neural networks accelerator that considers parallel computing efficiency. In addition, the proposed method proposes Optimal-FSRCNN, which is modified the structure of FSRCNN. The number of multipliers is compressed by 3.47 times compared to FSRCNN. Moreover, PSNR has similar performance to FSRCNN. We developed a real-time image processing technology that implements on FPGA.