• Proceedings of the KAIS Fall Conference
• /
• 2008.05a
• /
• pp.286-289
• /
• 2008

홍채인식은 홍채의 무늬 패턴 정보를 이용하여 동일인 여부를 판별하는 생체인식 기술이다. 최근 들어 홍채정보를 이용하여 출입통제, 정보보안등의 분야에 많이 활용되고 있다. 이러한 홍채 인식 시스템에 있어 조명의 영향이나 동공의 크기, 홍채의 회전 등 홍채 취득시 다양한 환경 조건으로 인해 발생될 수 있는 홍채 무늬 패턴의 변화에 강인한 특징을 추출하는 것은 홍채인식에 있어 매우 중요한 과제이다. 본 논문에서는 국부적 방향 히스토그램을 이용한 새로운 홍채 특징 추출 방법을 제안하였다. 제안된 방법은 성능면에서 기존의 방법들과 비교하여 대등한 성능을 보여주는 것을 실험을 통해 확인하였다.

• Journal of Broadcast Engineering
• /
• v.17 no.2
• /
• pp.316-324
• /
• 2012

In this paper, a novel descreening method using HOG(histogram of gradient)-based adaptive smoothing filter is proposed. Conventional edge-oriented smoothing methods does not provide enough smoothing to the halftone image due to the edge-like characteristic of the halftone noise. Moreover, clustered-dot halftoning method, which is commonly used in printing tends to create Moire pattern because of the intereference in color channels. Therefore, the proposed method uses HOG to distinguish edges and the amount of smoothing to be performed on the halftone image is then calculated according to the magnitude of the HOG in the edge and edge normal orientation. The proposed method was tested on various scanned halftone materials, and the results show that it effectively removes halftone noises as well as Moire pattern while preserving image details.

• Journal of Korea Multimedia Society
• /
• v.17 no.3
• /
• pp.300-311
• /
• 2014

According to fast change of the environment, the structured study of the ecosystem by analyzing the plant leaves are needed. Expecially, the methodology that searches and classifies the leaves from captured from the smart device have received numerous concerns in the field of computer science and ecology. In this paper, we propose a plant leaf classification technique using shape descriptor by combining Scale Invarinat Feature Transform (SIFT) and Histogram of Oriented Gradient (HOG) from the image segmented from the background via Graphcut algorithm. The shape descriptor is coded in the field of Locality-constrained Linear Coding to optimize the meaningful features from a high degree of freedom. It is connected to Support Vector Machines (SVM) for efficient classification. The experimental results show that our proposed approach is very efficient to classify the leaves which have similar color, and shape.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.10
• /
• pp.27-34
• /
• 2010

This paper proposes an efficient method for verifying palmprint which is captured at the natural interface without any physical restriction. The location and orientation of the region of interest (ROI) in palm images are variously appeared due to the translation and rotation of hand. Therefore, it is necessary to extract the ROI stably for palmprint recognition. This paper presents a method that can extract the ROI, which is based on the reference points that are located at the center of the crotch segments between index finger and middle finger and between ring finger and little finger. It also proposes a palmprint recognition method using the histogram of local binary patterns (LBP). Experiments for evaluating the performance of the proposed method were performed on 1,597 palmprint images acquired from 100 different persons. The experimental results showed that ROI was correctly extracted at the rate of 99.5% and the equal error rate (EER) and the decidability index d' indicating the performance of palmprint verification were 0.136 and 3.539, respectively. These results demonstrate that the proposed method is robust to the variations of the translation and rotation of hand.

• Journal of Intelligence and Information Systems
• /
• v.8 no.1
• /
• pp.27-39
• /
• 2002

A locating text technique capable of locating and extracting text blocks in various Web images is presented here. Until now this area of work has been ignored by researchers even if this sort of text may be meaningful for internet users. The algorithms associated with the technique work without prior knowledge of the text orientation, size or font. In the work presented in this research, our text extraction algorithm utilizes useful edge detection followed by histogram analysis on the genuine characteristics of letters defined by text clustering region, to properly perform extraction of the text region that does not depend on font styles and sizes. By a number of experiments we have showed impressively acceptable results.

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.27 no.1
• /
• pp.17-24
• /
• 1997

The purpose of the present study was to use digital profile image features and digital image analysis of fixed-dimension bone regions, extracted from standardized periapical radiographs of the maxilla, to determine whether differences exist in alveolar bone of younger women(mean age: 59.23±7.34 years) and just menopaused women(mean age: 59.23±7.34). Periapical films were used from two groups of 20 randomly selected women. None of the subjects had a remarkable medical history. To simplify protocol, we chose one interproximal bone area between the maxillary right canine and lateral incisor for study. Ech film was digitized into a 1312 x 1024 pixel x 8 bit depth matrix by means of a Nikon 35 mm film scanner(LS-35lOAF, Japan) with fixed gain and internal dark current correction to maintain constant illumination. The scanner was interfaced to a Macintosh LC III computer(Apple Computer, Charlotte, N.C.). Area and profile orientation were selected with a NIMH Image 1.37(NIH Research Services Branch, Bethesda, Md.). Histogram features were extracted from each profile and area. The results of this study indicate that mean pixel intensities didn't differ significantly between two groups and there was a high correlarion-coefficient between digitized radiographic profile features and area features.

• Journal of Institute of Control, Robotics and Systems
• /
• v.11 no.5
• /
• pp.423-428
• /
• 2005

This paper proposes a new edge detection method using a $3{\times}3$ ideal binary pattern and lookup table (LUT) for the mobile robot localization without any parameter adjustments. We take the mean of the pixels within the $3{\times}3$ block as a threshold by which the pixels are divided into two groups. The edge magnitude and orientation are calculated by taking the difference of average intensities of the two groups and by searching directional code in the LUT, respectively. And also the input image is not only partitioned into multiple groups according to their intensity similarities by the histogram, but also the threshold of each group is determined by fuzzy reasoning automatically. Finally, the edges are determined through non-maximum suppression using edge confidence measure and edge linking. Applying this edge detection method to the mobile robot localization using projective invariance of the cross ratio. we demonstrate the robustness of the proposed method to the illumination changes in a corridor environment.

• International Journal of Automotive Technology
• /
• v.6 no.2
• /
• pp.171-181
• /
• 2005

The purpose of this paper is two-fold: 1) A novel algorithm in order to extract lane-related information from road images is presented; 2) Design specifications of an image processing onboard unit capable of extracting lane­related information in real-time is also presented. Obtaining precise information from road images requires many features due to the effects of noise that eventually leads to long processing time. By exploiting a FPGA and DSP, we solve the problem of real-time processing. Due to the fact that image processing of road images relies largely on edge features, the FPGA is adopted in the hardware design. The schematic configuration of the FPGA is optimized in order to perform 3 $\times$ 3 Sobel edge extraction. The DSP carries out high-level image processing of recognition, decision, estimation, etc. The proposed algorithm uses edge features to define an Edge Distribution Function (EDF), which is a histogram of edge magnitude with respect to the edge orientation angle. The EDF enables the edge-related information and lane-related to be connected. The performance of the proposed system is verified through the extraction of lane-related information. The experimental results show the robustness of the proposed algorithm and a processing speed of more than 25 frames per second, which is considered quite successful.

• The Journal of the Korea Contents Association
• /
• v.10 no.8
• /
• pp.11-17
• /
• 2010

In this paper, we propose HMM-based upper-body gesture. First, to recognize gesture of space, division about pose that is composing gesture once should be put priority. In order to divide poses which using interface, we used two IR cameras established on front side and side. So we can divide and acquire in front side pose and side pose about one pose in each IR camera. We divided the acquired IR pose image using SVM's non-linear RBF kernel function. If we use RBF kernel, we can divide misclassification between non-linear classification poses. Like this, sequences of divided poses is recognized by gesture using HMM's state transition matrix. The recognized gesture can apply to existent application to do mapping to OS Value.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.1
• /
• pp.67-78
• /
• 2005

We proposed a blind watermarking for 3D mesh model using the patch CEGIs. The CEGI is the 3D orientation histogram with complex weight whose magnitude is the mesh area and phase is the normal distance of the mesh from the designated origin. In the proposed algorithm we divide the 3D mesh model into the number of patch that determined adaptively to the shape of model and calculate the patch CEGIs. Some cells for embedding the watermark are selected according to the rank of their magnitudes in each of patches after calculating the respective magnitude distributions of CEGI for each patches of a mesh model. Each of the watermark bit is embedded into cells with the same rank in these patch CEGI. Based on the patch center point and the rank table as watermark key, watermark extraction and realignment process are performed without the original mesh. In the rotated model, we perform the realignment process using Euler angle before the watermark extracting. The results of experiment verify that the proposed algorithm is imperceptible and robust against geometrical attacks of cropping, affine transformation and vertex randomization as well as topological attacks of remeshing and mesh simplification.