• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.3
• /
• pp.339-349
• /
• 2017

This paper proposes a four-stage algorithm for detecting lanes on a driving car. In the first stage, it extracts region of interests in an image. In the second stage, it employs a median filter to remove noise. In the third stage, a binary algorithm is used to classify two classes of backgrond and foreground of an input image. Finally, an image erosion algorithm is utilized to obtain clear lanes by removing noises and edges remained after the binary process. However, the proposed lane detection algorithm requires high computational time. To address this issue, this paper presents a parallel implementation of a real-time line detection algorithm on a multi-core architecture. In addition, we implement and simulate 8 different processing element (PE) architectures to select an optimal PE architecture for the target application. Experimental results indicate that 40×40 PE architecture show the best performance, energy efficiency and area efficiency.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.20 no.4
• /
• pp.39-54
• /
• 2017

Between 1998 and 2010, the computerization of underground facilities was completed in 84 cities. Since 2011, new pipelines have been laid or existing pipelines have been maintained, renovated, and renewed. To measure the exact location and depth of the exposure pipe, a map of underground facilities was created before filling the ground. This method is based on the time when the underground facilities of the National Geographic Information Institute Regulation No. 134 of the National Geographic Information Office revised in 2010 were drafted. The process of the drone taking the video is based on a theoretical basis of ground control points. The method works by removing all ground control points located outside of the error range and re-processing it for calculating the best result. Furthermore, using a drone-based spontaneous measuring method allows workers to obtain a high accuracy underground facilities map in error bound. The proposed method could be used as a new way to standardize the processing.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.3
• /
• pp.51-57
• /
• 2021

In this paper, we propose a method to find the DoF(Depth of field) that is blurred in an image by focusing and out-focusing the camera through a efficient convolutional neural network. Our approach uses the RGB channel-based cross-correlation filter to efficiently classify the DoF region from the image and build data for learning in the convolutional neural network. A data pair of the training data is established between the image and the DoF weighted map. Data used for learning uses DoF weight maps extracted by cross-correlation filters, and uses the result of applying the smoothing process to increase the convergence rate in the network learning stage. The DoF weighted image obtained as the test result stably finds the DoF region in the input image. As a result, the proposed method can be used in various places such as NPR(Non-photorealistic rendering) rendering and object detection by using the DoF area as the user's ROI(Region of interest).

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06a
• /
• pp.450-453
• /
• 2011

고화질의 디지털 카메라 및 스마트폰, 감시용 카메라의 보급 등으로 인해 최근 패턴 인식 및 이미지 프로세싱 분야에서 고화질의 이미지 및 비디오를 처리해야 하는 경우가 많아지고 있다. 특히 차량 번호판 감지 등과 같은 객체 인식 분야의 경우, 고화질의 이미지로 인해 그만큼 인식에 필요한 계산 비용이 증가하게 되었는데 따라서 이러한 계산 비용을 효율적으로 줄이기 위한 기법이 요구되고 있다. 또한 기존의 차량 번호판 감지의 도메인과는 다르게 도로 상에서의 실시간 차량 번호판 감지의 필요성이 대두되고 있기에 본 논문에서는 도로 상에서의 실시간 번호판 감지 시스템을 위한 차량 번호판 주변정보 기반의 효율적인 윈도우 슬라이딩(window sliding) 방법을 제안한다. 본 논문의 시스템은 총 3단계로, (1) SVM(Supported Vector Machine) 을 통한 차량 번호판 주위 정보에 대한 학습, (2) 도로 상의 번호판 위치 확률 모델링을 통한 탐색 공간의 감소, (3) $context_{plate}$분류기를 통한 OCS(operator context scanning)의 수행이다. 이와 같은 $context_{plate}$분류기와 OCS를 통해 번호판 검출을 위한 윈도우 슬라이딩의 수가 크게 줄었음을 알 수 있었으며, 또한 번호판의 정보를 건너뛰지 않고, 신뢰성 있게 접근함을 알 수 있었다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.5 no.1
• /
• pp.69-72
• /
• 2004

Compared to the barcode which is being widely employed to store information on products, the color code may find more various applications because of its favorable appearance and larger possible number of combinations. However, the color values read in practice may suffer from distortions from environments and devices. In this paper, we propose efficient ways to reduce the effect of such distortions and to detect color codes. for which we apply the Right Hand on Wall (RHOW) algorithm originated from the area of the maze search. The color codes used in this paper have high values of Hue and Saturation components and have a circular shape. We first preprocessed the images to detect candidate areas of color codes, and then applied the RHOW algorithm to determine optimal coordinates of rectangles enclosing the areas. As a result, we could obtain accurate coordinates of color codes by using the RHOW algorithm.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.1153-1156
• /
• 2005

PACS system developments for improve of diagnosis efficiency for hospital duty efficiency and patients through various medical image diagnosis device. PACS system studies for nonfilm hospital, but still, image filtering technology Present high accuracy take a photograph image be not enough. On study image processing part is very importance. The fittest of image processing in order to using simple and essential filter try to improvrnent. The future, diagnosis of patients decide accuracy, low cost based Digital XRay system will be using.

• Journal of Digital Contents Society
• /
• v.19 no.6
• /
• pp.1177-1183
• /
• 2018

Recent developments in machine learning have attracted a lot of attention for techniques such as machine learning and deep learning that implement artificial intelligence. In this paper, binary malicious code using deep learning based R-CNN is imaged and the feature is extracted from the image to classify the family. In this paper, two steps are used in deep learning to image malicious code using CNN. And classify the characteristics of the family of malicious codes using R-CNN. Generate malicious code as an image, extract features, classify the family, and automatically classify the evolution of malicious code. The detection rate of the proposed method is 93.4% and the accuracy is 98.6%. In addition, the CNN processing speed for image processing of malicious code is 23.3 ms, and the R-CNN processing speed is 4ms to classify one sample.

• Journal of Korea Multimedia Society
• /
• v.5 no.3
• /
• pp.281-289
• /
• 2002

Few studies have been systematically pursued on a multimedia data mining in despite of the over-whelming amounts of multimedia data by the development of computer capacity, storage technology and Internet. Based on the preliminary image processing and content-based image retrieval technology, this paper presents the methods for discovering association rules from recurrent items with spatial relationships in huge data repositories. Furthermore, multimedia mining algorithm is proposed to find implicit association rules among objects of which content-based descriptors such as color, texture, shape and etc. are recurrent and of which descriptors have spatial relationships. The algorithm with recurrent items in images shows high efficiency to find set of frequent items as compared to the Apriori algorithm. The multimedia association-rules algorithm is specially effective when the collection of images is homogeneous and it can be applied to many multimedia-related application fields.

• The KIPS Transactions:PartD
• /
• v.9D no.1
• /
• pp.57-64
• /
• 2002

Few studies have been systematically pursued on a multimedia data mining in despite of the overwhelming amounts of multimedia data by the development of computer capacity, storage technology and Internet. Based on the preliminary image processing and content-based image retrieval technology, this paper presents the methods for discovering association rules from recurrent items with spatial relationships in huge data repositories. Furthermore, multimedia mining algorithm is proposed to find implicit association rules among objects of which content-based descriptors such as color, texture, shape and etc. are recurrent and of which descriptors have spatial relationships. The algorithm with recurrent items in images shows high efficiency to find set of frequent items as compared to the Apriori algorithm. The multimedia association-rules algorithm is specially effective when the collection of images is homogeneous and it can be applied to many multimedia-related application fields.

• Journal of the Korean Geotechnical Society
• /
• v.27 no.11
• /
• pp.83-92
• /
• 2011

The random and heterogeneous pore structure is a significant factor that dominates physical and mechanical behaviors of soils such as fluid flow and geomechanical responses driven by loading. The characterization method using non-destructive testing such as micro X-ray CT technique which has a high resolution with micrometer unit allows to observe internal structure of soils. However, the application has been limited to qualitatively observe 2D and 3D CT images and to obtain the void ratio at macro-scale although the CT images contain enormous information of materials of interests. In this study, we constructed the 3D particle and pore structures based on sequentially taken 2D images of glass beads and quantitatively defined complex pore structure with void cell and void channel. This approach was enabled by implementing image processing techniques that include coordinate transformation, binarization, Delaunay Triangulation, and Euclidean Distance Transform. It was confirmed that the suggested algorithm allows to quantitatively evaluate the distribution of void cells and their connectivity of heterogeneous pore structures for glass beads.