• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.4
• /
• pp.125-132
• /
• 2021

As the layers of artificial neural network deepens, and the dimension of data used as an input increases, there is a problem of high arithmetic operation requiring a lot of arithmetic operation at a high speed in the learning and recognition of the neural network (NN). Thus, this study proposes a data dimensionality reduction method to reduce the dimension of the input data in the NN. The proposed Line-segment Feature Analysis (LFA) algorithm applies a gradient-based edge detection algorithm using median filters to analyze the line-segment features of the objects existing in an image. Concerning the extracted edge image, the eigenvalues corresponding to eight kinds of line-segment are calculated, using 3×3 or 5×5-sized detection filters consisting of the coefficient values, including [0, 1, 2, 4, 8, 16, 32, 64, and 128]. Two one-dimensional 256-sized data are produced, accumulating the same response values from the eigenvalue calculated with each detection filter, and the two data elements are added up. Two LFA256 data are merged to produce 512-sized LAF512 data. For the performance evaluation of the proposed LFA algorithm to reduce the data dimension for the recognition of handwritten numbers, as a result of a comparative experiment, using the PCA technique and AlexNet model, LFA256 and LFA512 showed a recognition performance respectively of 98.7% and 99%.

• Journal of Korea Game Society
• /
• v.6 no.2
• /
• pp.3-12
• /
• 2006

Recently, Many 3D augmented reality games that provide strengthened immersive have appeared in the 3D game environment. In this article, we describe a barehanded interaction method based on human hand gestures for augmented reality games. First, feature points are extracted from input video streams. Point features are tracked and motion of moving objects are computed. The shape of the motion trajectories are used to determine whether the motion is intended gestures. A long smooth trajectory toward one of virtual objects or menus is classified as an intended gesture and the corresponding action is invoked. To prove the validity of the proposed method, we implemented two simple augmented reality applications: a gesture-based music player and a virtual basketball game. In the music player, several menu icons are displayed on the top of the screen and an user can activate a menu by hand gestures. In the virtual basketball game, a virtual ball is bouncing in a virtual cube space and the real video stream is shown in the background. An user can hit the virtual ball with his hand gestures. From the experiments for three untrained users, it is shown that the accuracy of menu activation according to the intended gestures is 94% for normal speed gestures and 84% for fast and abrupt gestures.

• The KIPS Transactions:PartB
• /
• v.15B no.1
• /
• pp.25-36
• /
• 2008

In this paper, we present a robust hand recognition approach to sudden illumination changes. The proposed approach constructs a background model with respect to hue and hue gradient in HSI color space and extracts a foreground hand region from an input image using the background subtraction method. Eighteen features are defined for a hand pose and multi-class SVM(Support Vector Machine) approach is applied to learn and classify hand poses based on eighteen features. The proposed approach robustly extracts the contour of a hand with variations in illumination by applying the hue gradient into the background subtraction. A hand pose is defined by two Eigen values which are normalized by the size of OBB(Object-Oriented Bounding Box), and sixteen feature values which represent the number of hand contour points included in each subrange of OBB. We compared the RGB-based background subtraction, hue-based background subtraction and the proposed approach with sudden illumination changes and proved the robustness of the proposed approach. In the experiment, we built a hand pose training model from 2,700 sample hand images of six subjects which represent nine numerical numbers from one to nine. Our implementation result shows 92.6% of successful recognition rate for 1,620 hand images with various lighting condition using the training model.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.33-41
• /
• 2007

Owing to the development of multimedia and communication technologies, multimedia data become a common feature of the information systems and are on the increase. This has led to the need of 3D shape retrieval systems that, given a query object, retrieve similar 3D objects. Therefore, shape descriptor required to describe a 3D object effectively and efficiently. In this paper, a new descriptor for 3D model retrieval based on shape information is proposed. The proposed descriptor utilizes the curvedness together with the shape index that provides local geometry information. The existing 3D Shape Spectrum Descriptor (3D SSD), which is defined as the histogram of shape index values, represents the characteristics of local shapes of the 3D surface. However, it does not properly represent the local shape characteristics, because many points with different curvedness may have the same shape index value. Therefore, we add a new feature that represents the degree of curvedness, thereby improving the discriminating power of the shape descriptor. We evaluate the performance of the proposed method, compared with the previous method. The experimental results have shown that the performance of retrieval has been improved by 23.6%.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.30 no.1
• /
• pp.39-48
• /
• 2012

Filtering process that separates ground and non-ground points from LIDAR data is important in order to create the digital elevation model (DEM) or extract objects on the ground. The purpose of this research is to select the most effective filtering algorithm through qualitative and quantitative analysis for the existing filtering method used to extract ground points from LIDAR data. For this, four filtering methods including Adaptive TIN(ATIN), Perspective Center-based filtering method(PC), Elevation Threshold with Expand Window(ETEW) and Progressive Morphology(PM) were applied to mountain area, urban area and the area where building and mountains exist together. Then the characteristics for each method were analyzed. For the qualitative comparison of four filtering methods used for the research, visual method was applied after creating shaded relief image. For the quantitative comparison, an absolute comparison was conducted by using control points observed by GPS and a relative comparison was conducted by the digital elevation model of the National Geographic Information Institute. Through the filtering experiment of the LIDAR data, the Adaptive TIN algorithm extracted the ground points in mountain area and urban area most effectively. In the area where buildings and mountains coexist, progressive morphology algorithm generated the best result. In addition, as a result of qualitative and quantitative comparisons, the applicable filtering algorithm regardless of topographic characteristics appeared to be ATIN algorithm.

• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.612-628
• /
• 2001

This paper presents a methodology to extract classes and inheritance relations from procedural software. The methodology is based on the idea of generating all groups of class candidates, based on the combinatorial groups of object candidates, and their inheritance with all possible combinations and selecting a group of object candidates, and their inheritance with all possible combinations and selecting a group with the best or optimal combination of candidates with respect to the degree of relativity and similarity between class candidates in the group and classes in a domain model. The methodology has innovative features in class candidates in the group and classes in a domain model. The methodology has innovative features in class and inheritance extraction: a clustering method based on both static (attribute) and dynamic (method) clustering, the combinatorial cases of grouping class candidate cases based on abstraction, a signature similarity measurement for inheritance relations among n class candidates or m classes, two-dimensional similarity measurement for inheritance relations among n class candidates or m classes, two-dimensional similarity measurement, that is, the horizontal measurement for overall group similarity between n class candidates and m classes, and the vertical measurement for specific similarity between a set of classes in a group of class candidates and a set of classes with the same class hierarchy in a domain model, etc. This methodology provides reengineering experts with a comprehensive and integrated environment to select the best or optimal group of class candidates.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.12
• /
• pp.113-123
• /
• 2012

Recently, various attempts have been made to apply Intelligent Transportation System under various environments and conditions. Consequently, an accurate license plate recognition regardless of vehicle location and viewing angle is required. In this paper, we propose a novel license plate recognition system which exploits a) the format of license plates to remove false candidates of license plates and to extract characters in license plates and b) the characteristics of Hangul for accurate character recognition. In order to eliminate false candidates of license plates, the proposed method first aligns the candidates of license plates horizontally, and compares the position and the shape of objects in each candidate with the prior information of license plates provided by Korean Ministry of Construction & Transportation. The prior information such as aspect ratio, background color, projection image is also used to extract characters in license plates accurately applying an improved local binarization considering luminance variation of license plates. In case of recognizing Hangul in license plates, they are initially grouped according to their shape similarity. Then a super-class method, a hierarchical analysis based on key feature points is applied to recognize Hangul accurately. The proposed method was verified with high recognition rate regardless of background image, which eventually proves that the proposed LPR system has high performance regardless of the vehicle location or viewing angle.

• The KIPS Transactions:PartC
• /
• v.10C no.5
• /
• pp.647-656
• /
• 2003

In Webspace, mining traversal patterns is to understand user's path traversal patterns. On this mining, it has a unique characteristic which objects (for example, URLs) may be visited due to their positions rather than contents, because users move to other objects according to providing information services. As a consequence, it becomes very complex to extract meaningful information from these data. Recently discovering traversal patterns has been an important problem in data mining because there has been an increasing amount of research activity on various aspects of improving the quality of information services. This paper presents a Dynamic Link Recommendation (DLR) algorithm that recommends link sets on a Web site through mining frequent traversal patterns. It can be employed to any Web site with massive amounts of data. Our experimentation with two real Weblog data clearly validate that our method outperforms traditional method.

• Journal of Digital Convergence
• /
• v.11 no.2
• /
• pp.215-220
• /
• 2013

They are well-known geometric properties of a circle that the perpendicular bisector of a chord passes through the center of a circle, and the intersection of the perpendicular bisectors of any two chords is its center. This paper is related to a fast pupil detection method capable of detecting the center and the radius of a pupil using these geometric properties at high speed when detecting the pupil region for iris segmentation. The proposed method is characterized as rapidly detecting the center and the radius of the pupil, extracting the candidate points of the circle in human eye images using morphological operations, and finding two chords using four points on the circular edge, and taking the intersection of the perpendicular bisectors of these two chords for its center. The proposed method can not only detect the center and the radius of a pupil rapidly but also find partially occluded pupils in human eye images.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.4
• /
• pp.127-133
• /
• 2009

To ensure access to rapidly growing video collection, video indexing is becoming more and more essential. A database for video should be build for fast searching and extracting the accurate features of video information with more complex characteristics. Moreover, video indexing structure supports efficient retrieval of interesting contents to reflect user preferences. In this paper, we propose semantic video retrieval method based on user preference. Unlikely the previous methods do not consider user preferences. Futhermore, the conventional methods show the result as simple text matching for the user's query that does not supports the semantic search. To overcome these limitations, we develop a method for user preference analysis and present a method of video ontology construction for semantic retrieval. The simulation results show that the proposed algorithm performs better than previous methods in terms of semantic video retrieval based on user preferences.