• Journal of Korea Multimedia Society
• /
• v.14 no.7
• /
• pp.858-865
• /
• 2011

In order to allow the user to efficiently browse, select, and retrieve a desired video part without having to deal directly with GBytes of compressed data, classification of shot motion characteristic has to be carried out as a preparation for such user interaction. The organization of video information for video database requires segmentation of a video into its constituent shots and their subsequent characterization in terms of content and camera movement in shot. In order to classify shot motion, it is a conventional way to use element of motion vector. However, there is a limit to estimate global camera motion because the way that uses motion vectors only represents local movement. For shot classification in terms of motion information, we propose a new scheme consisting of partial decoding of INTRA pictures and comparing the x, y displacement vector curve between the decoded I-frame and next P-frame in compressed video data.

• Journal of Advanced Navigation Technology
• /
• v.18 no.2
• /
• pp.151-157
• /
• 2014

Recently UAV(unmanned aerial vehicle) is frequently used not only for military purpose but also for civil purpose. UAV automatically navigates following the coordinates input in advance using GPS information. However it is impossible when GPS cannot be received because of jamming or external interference. In order to solve this problem, we propose a real-time segmentation and classification algorithm for the specific regions from UAV image in this paper. We use the super-pixels algorithm using graph-based image segmentation as a pre-processing stage for the feature extraction. We choose the most ideal model by analyzing various color models and mixture color models. Also, we use support vector machine for classification, which is one of the machine learning algorithms and can use small quantity of training data. 18 color and texture feature vectors are extracted from the UAV image, then 3 classes of regions; river, vinyl house, rice filed are classified in real-time through training and prediction processes.

• The Journal of the Korea Contents Association
• /
• v.13 no.8
• /
• pp.1-10
• /
• 2013

This paper proposes a method to detect vanishing points of an indoor scene using line segment classification. Two-stage vanishing points detection is carried out to detect vanishing point in indoor scene efficiently. In the first stage, the method examines whether the image composition is a one-point perspective projection or a two-point one. If it is a two-point perspective projection, a horizontal line through the detected vanishing point is found for line segment classification. In the second stage, the method detects two vanishing points exactly using line segment classification. The method is evaluated by synthetic images and an image DB. In the synthetic image which some noise is added in, vanishing point detection error is under 16 pixels until the percent of the noise to the image becomes 60%. Vanishing points detection ratio by A.Quattoni and A.Torralba's image DB is over 87%.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.4
• /
• pp.57-67
• /
• 2015

The concept of detection and classification of objects based on infrared camera is widely applied to military applications. While the object detection technology using infrared images has long been researched and the latest one can detect the object in sub-pixel, the object classification technology still needs more research. In this paper, we present object classification method based on measured radiant intensity of objects such as target, artillery, and missile using infrared camera. The suggested classification method was verified by radiant intensity measuring experiment using black body. Also, possible measuring errors were compensated by modelling-based correction for accurate radiant intensity measure. After measuring radiation of object, the model of radiant intensity is standardized based on theoretical background. Based on this research, the standardized model can be applied to the object classification by comparing with the actual measured radiant intensity of target, artillery, and missile.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.17 no.2
• /
• pp.121-135
• /
• 2014

Object-based classification methods provide a valid alternative to traditional pixel-based methods. This study reports the results of an object-based classification to examine land cover in the demilitarized zones(DMZs) of Inje-gun. We used land cover classes(7 classes for main category and 13 classes for sub-category) selected from the criteria by Korea Ministry of Environment. The average and standard deviation of the spectrum values, and homogeneity of GLCM were chosen to map land cover types in an hierarchical approach using the nearest neighborhood method. We then identified the distributional characteristics of land cover by considering 3 topographic characteristics (altitude, slope gradient, distance from the Southern Limited Line(SLL)) within the DMZs. The results showed that scale 72, shape 0.2, color 0.8, compactness 0.5 and smoothness 0.5 were the optimum weight values while scale, shape and color were most influenced parameters in image segmentation. The forests (92%) were main land cover type in the DMZs; the grassland(5%), the urban area (2%) and the forests (broadleaf forest: 44%, mixed forest: 42%, coniferous forest: 6%) also occupied mostly in land cover classes for sub-category. The results also showed that facilities and roads had higher density within 2 km from the SLL, while paddy, field and bare land were distributed largely outside 6 km from the SLL. In addition, there was apparent distinction in land cover by topographic characteristics. The forest had higher density at above altitude 600m and above slope gradient $30^{\circ}$ while agriculture, bare land and grass land were distributed mainly at below altitude 600m and below slope gradient $30^{\circ}$.

• Journal of Korea Multimedia Society
• /
• v.6 no.5
• /
• pp.778-787
• /
• 2003

Intrusion detection systems based on rules are not efficient for mutated attacks, because they need additional rules for the variations. In this paper, we propose an intrusion detection system using the time delay neural network. Packets on the network can be considered as gray images of which pixels represent bytes of them. Using this continuous packet images, we construct a neural network classifier that discriminates between normal and abnormal packet flows. The system deals well with various mutated attacks, as well as well known attacks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.10C
• /
• pp.1000-1009
• /
• 2007

This paper presents a content adaptive interpolation (CAI) for intra deinterlacing. The CAI consists of three steps: pre-processing, content classification, and adaptive interpolation. There are also three main interpolation methods in our proposed CAI, i.e. modified edge-based line averaging (M-ELA), gradient directed interpolation (GDI), and window matching method (WMM). Each proposed method shows different performances according to spatial local features. Therefore, we analyze the local region feature using the gradient detection and classify each missing pixel into four categories. And then, based on the classification result, a different do-interlacing algorithm is activated in order to obtain the best performance. Experimental results demonstrate that the CAI method performs better than previous techniques.

• Journal of Biomedical Engineering Research
• /
• v.19 no.1
• /
• pp.1-8
• /
• 1998

Clustered microcalcifications on X-ray mammograms are an important sign for early detection of breast cancer. This paper proposes a computer-aided diagnosis method for the detection of clustered microcalcifications and marking their locations on digitized mammograms. The proposed detection method consists of the region of interest (ROI) selection, the film-artifact removal, the surrounding texture analysis method for the detection of clustered microcalcifications, which is based on the second-order histogram in two nested surrounding regions on the current pixel. This paper also describes the effectiveness of the proposed film-artifact removal filter in terms of the classification performance with the receiver operating-characteristics(ROC) analysis. A three-layer backpropagation neural network is employed as a classifier. The appropriate marking for the locations of clustered microcalcifications can be used to alert radiologists to locations of suspicious lesions.

• Journal of Korea Multimedia Society
• /
• v.6 no.1
• /
• pp.78-88
• /
• 2003

A novel adaptive error concealment technique is proposed using a variable operating region algorithm based on MPEG-4 coding. In the algorithm, a missing block is classified as flat or edge block based on local information from the surrounding blocks extracted using a Sobel operation in a variable operating region (VOR). In this case, the VOR is determined adaptively according to the number of edge directions in the missing block. 1;sing the classification, the flat blocks are then concealed by the Proposed mean based weighted bilinear interpolation (MWBLI) method, and the edge blocks by the boundary directional interpolation (BDI) method. Consequently, the use of the Proposed VOR improves the subjective performance in a curved edge region, while the adaptive processing based on block classification improves the objective performance. Experimental results confirmed that the proposed algorithm produced better results than conventional algorithms, both subjectively and objectively.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.17 no.4
• /
• pp.189-199
• /
• 2014

Floodplain is a flat plain between levees and rivers. This paper suggests a methodology for analyzing the land uses in the Nakdong River floodplain using the RapidEye imagery and the given LiDAR(LIght Detection And Ranging) DEM(Digital Elevation Models). First, the levee boundaries are generated using the LiDAR DEM, and the area of the floodplain is extracted from the given RapidEye imagery. The land uses in the floodplain are identified in the extracted RapidEye imagery by the ISODATA(Iterative Self-Organizing Data Analysis Technique Analysis) clustering. The overall accuracy of the identified land uses by the ISODATA clustering is 91%. Analysis of the identified land uses in the floodplain is implemented by counting the number of the pixels constituting the land cover clusters. The results of this research shows that the area of the river occupies 46%, the area of the bare soil occupies 36%, the area of the marsh occupies 11%, and the area of the grass occupies 7% in the identified floodplain.