Seok-Woo Jang

Analysis of Digital Hologram Rendering Using a Computational Method

To manufacture a real time digital holographic display system capable of being applied to next-generation television, it is important to rapidly generate a digital hologram. In this paper, we analyze digital hologram rendering based on a computer computation scheme. We analyze previous recursive methods to identify regularity between the depth-map image and the digital hologram.

Texture feature-based text region segmentation in social multimedia data

This paper proposes a method of effectively segmenting text areas that exist in images by using the texture features of various types of input images obtained in social multimedia networks with an artificial neural network. The proposed text segmentation method consists of four main steps: a step for extracting candidate text areas, a step for localizing the text areas, a step for separating the text from the background, and a step for verifying the candidate text areas. In the candidate text area extraction step, candidate blocks that have any text areas are segmented in an input image on the basis of the texture features of the candidate blocks. In the text area localization step, only strings are extracted from the candidate text blocks. In the text and background separation step, the text areas are separated from the background area in the localized text blocks. In the candidate text area verification step, an artificial neural network is used to verify whether the extracted text blocks include actual text areas and exclude non-text areas. In the experimental results, the proposed method was applied to various types of news and non-news images, and it was found that the proposed method extracted text regions more accurately than existing methods.

ASM-Based Objectionable Image Detection in Social Network Services

This paper presents a method for detecting harmful images using an active shape model (ASM) in social network services (SNS). For this purpose, our method first learns the shape of a womans breast lines through principal component analysis and alignment, as well as the distribution of the intensity values of the corresponding control points. This method then finds actual breast lines with a learned shape and the pixel distribution. In this paper, to accurately select the initial positions of the ASM, we attempt to extract its parameter values for the scale, rotation, and translation. To obtain this information, we search for the location of the nipple areas and extract the location of the candidate breast lines by radiating in all directions from each nipple position. We then locate the mean shape of the ASM by finding the scale and rotation values with the extracted breast lines. Subsequently, we repeat the matching process of the ASM until saturation is reached. Finally, we determine objectionable images by calculating the average distance between each control point in a converged shape and a candidate breast line.

Detection of Harmful Content Using Multilevel Verification in Visual Sensor Data

Various types of harmful content such as adult images and video clips have been increasingly distributed through wired or wireless visual sensor-based networks. In this paper, we propose a new algorithm for extracting human nipple regions, representing the harmfulness of the images, by using a multilevel verification technique in visual sensor-based image data. The proposed algorithm first detects human face regions including eyes and lips from input images. The method then generates a nipple map utilizing representative features that female nipples have and detects candidate nipple regions by applying the generated nipple map to segmented skin regions followed by morphological operations. Subsequently, the proposed method selects real nipple areas after eliminating non-nipple regions at multiple levels by applying geometrical information and an average color filter to the detected candidate nipple regions. Experimental results show that the proposed method can robustly extract female nipple regions in various types of input images captured in environments where certain constraints are not imposed on.

An adaptive camera-selection algorithm to acquire higher-quality images

Various types of three-dimensional (3D) cameras have been used to analyze real-world objects or environments effectively. However, because most existing 3D cameras capture scenes by statically using one type of camera, there may be a limit to the quality of the captured images. Therefore, in this paper, we build a hybrid camera system that combines passive triangulation (PT)- and active triangulation (AT)-based cameras and suggest a new mechanism of estimating accurate 3D depth by adaptively switching between the two types of cameras depending on the complexity of the environment. The suggested method initially uses initial input images to extract brightness and texture, which are major features representing the current state of the surrounding environment. The method subsequently generates a set of rules that dynamically select the PT- or AT-based camera, whichever can operate more suitably in the current environment, by analyzing the two extracted features. In experimental results, we demonstrate that the proposed adaptive camera-selection approach can be applied to extract 3D depth reliably with reasonable performance in terms of accuracy and time.

Detection of Harmful Images Based on Color and Geometrical Features

Abstract Along with the development of high-speed, wired and wireless Internet technology, various harmful images in a form of photos and video clips have become prevalent these days. In this paper, we suggest a method of automatically detecting adult images by extracting womans nipple areas which represent obscenity of the image. The suggested algorithm first segments skin color areas in the YC b C r color space from input images and extracts nipples candidate areas from the segmented skin areas through the suggested nipple map. We then select real nipple areas by using geometrical information and determines input images as harmful images if they contain nipples. Experimental results show that the suggested nipple map-based method effectively detects adult images. Key Words : Color Space, Geometrical Feature, Harmful Image, Skin Color Region 이 논문은 2011년도 정부(교육과학기술부)의 재원으로 한국연구재단의 지원을 받아 수행된 기초연구사업임 (No. 2011-0021984) * Corresponding Author: Moon-Haeng Huh (Anyang University)Tel: +82-31-467-0803 email: moonh@anyang.ac.krReceived August 12, 2013 Revised September 2, 2013 Accepted November 7, 2013

Robust detection of mosaic regions in visual image data

Due to the explosive increase in the production and sharing of digital visual media such as photos, animations, films and video clips, the need to intentionally or unintentionally create mosaic blocks to effectively cover user-designated regions within images has grown. In this paper, a method using boundary characteristics to effectively detect the grid-type mosaic blocks existing within the input image is proposed. Initially, the Canny edge is detected from the input image. Then, the boundary characteristics of mosaic blocks are extracted from the detected edges, and the candidate regions which may contain mosaic blocks are detected. After this stage, geometric characteristics are used to eliminate non-mosaic blocks and select actual mosaic blocks. In the experiment performed in this paper, it was observed that the method using boundary characteristics achieved more robust detection of the grid-type mosaic blocks from various input images than other existing methods. The mosaic-detection method proposed in this paper is expected to be valuable for applications in various fields.

A Column-Aware Index Management Using Flash Memory for Read-Intensive Databases

Abstract Most traditional database systems exploit a record-oriented model where the attributes of a record are placed contiguously in a hard disk to achieve high performance writes. However, for read-mostly data warehouse systems, the column-oriented database has become a proper model because of its superior read performance. Today, flash memory is largely recognized as the preferred storage media for high-speed database systems. In this paper, we introduce a column-oriented database model based on flash memory and then propose a new column-aware flash indexing scheme for the high-speed column-oriented data warehouse systems. Our index management scheme, which uses an enhanced B + -Tree, achieves superior search performance by indexing an embedded segment and packing an unused space in internal and leaf nodes. Based on the performance results of two test databases, we concluded that the column-aware flash index management outperforms the traditional scheme in the respect of the mixed operation throughput and its response time.

Skin region segmentation using an image-adapted colour model

This paper presents a new skin region extraction method that generates an image-adapted skin colour model and then segments skin areas using the model. Our method first detects eyes by using an eye map and then develops an image-adapted skin colour distribution model based on a skin map generated by reliably selecting true skin samples near the detected eyes. All skin areas over the entire image are then segmented with the generated skin model. While most of the existing skin detection methods use some pre-defined colour model, our skin model is adaptively constructed from each test image online so that it can overcome fundamental difficulties in extracting skin areas. Experimental results show that our skin extraction method gives better results as compared to other existing approaches.

Target Object Detection Based on Robust Feature Extraction

Detecting target objects robustly in natural environments is a difficult problem in the computer vision and image processing areas. This paper suggests a method of robustly detecting target objects in the environments where reflection exists. The suggested algorithm first captures scenes with a stereo camera and extracts the line and corner features representing the target objects. This method then eliminates the reflected features among the extracted ones using a homographic transform. Subsequently, the method robustly detects the target objects by clustering only real features. The experimental results showed that the suggested algorithm effectively detects the target objects in reflection environments rather than existing algorithms.