• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.175-177
• /
• 2004

최근, 이미지 검색기법에서는 객체추출 방법이나 관심영역 추출방법에 관한 연구가 활발히 이루어지고 있다. 그러나, 컬러 이미지의 경우 색상을 고려한 관심영역 특징추출 방법이나 인덱스 기법은 많이 연구되지 못하고 있다. 따라서, 본 논문에서는 컬러 이미지의 색상을 기반으로 하는 사분트리 분할 인덱스 기법을 제안한다. 사분트리 분할 인덱스 구조는 컬러 이미지의 공간 영역을 계층적인 영역으로 분할하여 각 공간 영역의 평균 색상 갓을 데이터베이스에 저장한다 저장되어진 각 영역의 평균 색상은 검색의 효율성을 높이기 위해 사분트리 인스턴스(Quad-tree distance)를 퍼지 값으로 계산하여 인덱스를 생성한다. 생성된 사분트리 분할 인덱스는 컬러 이미지의 관심영역(Region of Interest)의 색상을 검색할 때 유용하게 사용되며. 검색속도의 향상에 도움을 준다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.299-302
• /
• 2023

입 모양 인식은 음성인식의 중요한 부분 중 하나로 이를 개선하기위한 다양한 연구가 진행되어 왔다. 기존의 연구에서는 주로 입술주변 영역을 관찰하고 인식하는데 초점을 두었으나, 본 논문은 음성인식 시스템에서 기존의 입술영역과 함께 입술, 턱, 뺨 등 다른 관심 영역을 고려하여 음성인식 시스템의 입모양 인식 성능을 비교하였다. 입 모양 인식의 관심 영역을 자동으로 검출하기 위해 객체 탐지 인공신경망을 사용하며, 이를 통해 다양한 관심영역을 실험하였다. 실험 결과 입술영역만 포함하는 ROI 에 대한 결과가 기존의 93.92%의 평균 인식률보다 높은 97.36%로 가장 높은 성능을 나타내었다.

• Journal of Korea Multimedia Society
• /
• v.7 no.3
• /
• pp.388-398
• /
• 2004

This paper focuses on the reduction method of message traffic occurred when multiple participants collaborate in the distributed virtual environment. It is referred to as filtering method of update message to minimize data traffic exchanged in the virtual environment, and it is the essential process to improve the scalability of the virtual environment. Spatial partitioning method, a popular filtering technique, divides the whole environment into smaller sub-parts in order to make it small the regions to be maintained consistently. However, it is less flexible and adaptable because the information of area of interest must be configured before simulation. In this paper, the concept of dynamic area of interest is proposed, which is highly adaptable by applying dynamic environmental elements to AOI management. It uses multiple layering and multiple consistency algorithms for dynamic management, and makes it possible to consume system resource efficiently. For performance evaluation, experiments are performed with virtual avatars, measuring message traffic. Finally, the proposed model could be applied to the AOI management systems which accommodate massive users like MMORPG, or virtual communities.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.493-497
• /
• 2016

Infrared (IR) imaging is capable of detecting targets that are not visible at night, thus it has been widely used for the security and defense system. However, the quality of the IR image is often degraded by low resolution and noise corruption. This paper addresses target segmentation with the IR image. Multiple regions of interest (ROI) are extracted by the multi-level segmentation and targets are segmented from the individual ROI. Each level of the multi-level segmentation is composed of a k-means clustering algorithm an expectation-maximization (EM) algorithm, and a decision process. The k-means clustering algorithm initializes the parameters of the Gaussian mixture model (GMM) and the EM algorithm iteratively estimates those parameters. Each pixel is assigned to one of clusters during the decision. This paper proposes the selection and the merging of the extracted ROIs. ROI regions are selectively merged in order to include the overlapped ROI windows. In the experiments, the proposed method is tested on an IR image capturing two pedestrians at night. The performance is compared with conventional methods showing that the proposed method outperforms others.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.10
• /
• pp.63-71
• /
• 2009

In this study, we propose a revised dynamic ROI (Region-of-Interest) coding method in which the focused ROI is automatically extracted without help from users during the recovery process of low DOF (Depth-of-Field) JPEG2000 image. The proposed method creates edge mask information using high frequency sub-band data on a specific level in DWT (Discrete Wavelet Transform), and then identifies the edge code block for a high-speed ROI extraction. The algorithm scans the edge mask data in four directions by the unit of code block and identifies the edge code block simply and fastly using a edge threshold. As the results of experimentation applying for Implicit method, the proposed method showed the superiority in the side of speed and quality comparing to the existing methods.

• Progress in Medical Physics
• /
• v.12 no.1
• /
• pp.41-50
• /
• 2001

The quantification analysis of renal scintigraphy is strongly affected by the location, shape and size of region of interest(ROI). When ROIs are drawn manually, these ROIs are not reproducible due to the operators' subjective point of view, and may lead to inconsistent results even if the same data were analyzed. In this study, the effect of the ROI variation on the analysis of renal scintigraphy when the ROIs are drawn manually was investigated, and in order to obtain more consistent results, methods for automated ROI definition were developed and the results from the application of the developed methods were analyzed. Relative renal function, glomerular filtration rate and mean transit time were selected as clinical parameters for the analysis of the effect of ROI and the analysis tools were designed with the programming language of IDL5.2. To obtain renal scintigraphy, $^{99m}$Tc-DTPA was injected to the 11 adults of normal condition and to study the inter-operator variability, 9 researchers executed the analyses. The calculation of threshold using the gradient value of pixels and border tracing technique were used to define renal ROI and then the background ROI and aorta ROI were defined automatically considering anatomical information and pixel value. The automatic methods to define renal ROI were classified to 4 groups according to the exclusion of operator's subjectiveness. These automatic methods reduced the inter-operator variability remarkably in comparison with manual method and proved the effective tool to obtain reasonable and consistent results in analyzing the renal scintigraphy quantitatively.

• Journal of KIISE:Software and Applications
• /
• v.27 no.8
• /
• pp.804-814
• /
• 2000

In this paper, we propose an effective method to detect the regions of interests in the Videostrobokymography System. Videostrobokymography system is a medical image processing system for extracting automatically the diagnosis parameters from the irregular vibratory movements of the vocal fold. We detect the regions of interests through three steps. In the first step, we remove the noise in the input image and we find the minimum energy value in each frame. In the second step, we computed the edge by everage value for the one line. In the third step, the regions of interests can be extracted by using the Merge Algorithm which uses the variance of luminance as the feature points. We experimented this method for the vocal fold images of nineteen patients. In consequence, the regions of interests are detected in most vocal fold images. The method proposed in this study is efficient enough to extract the region of interests in the vocal fold images with the frame rate of 40 frames/second and the resolution of 200${\times}$280 pixels.

• Progress in Medical Physics
• /
• v.14 no.4
• /
• pp.268-278
• /
• 2003

Purpose：The aim of this study was to evaluate the striatal binding ratio, the anterior/posterior ratio and reproducibility using a template based registration method using the standard MR template as a replacement for each patients MR image. Materials and Methods：This study analyzed the 123I IPT SPECT images of 30 patients with IPD, who were subdivided into 17 patients (56.6$\pm$10.8 yr, M/F : 8/9.) with mild IPD, and 13 patients (56.4$\pm$11.1 yr, M/F : 8/5) with severe IPD. In addition, 11 normal controls (57.8$\pm$14.4 yr, M/F : 4/7) were also analyzed. The ROIs were positioned manually in the same slice showing the highest striatal activity using the traditional manual method, whereas those were positioned automatically in a mid striatal slice of the SPECT image coregistered to the standard T1 weighted MR template. Results : The specific binding ratio (SBR) obtained using the template based registration method strongly correlated with those using the manual method in all groups : normal controls (r=0.85, P<0.001), mild IPD (r=0.84, P<0.001) and severe IPD (r=0.7, P=0.01). The SBRs obtained using both methods were significantly different among the three groups (P=0.05) and the SBRs obtained by the template based registration method were higher than those by the manual method (P=0.05) in all three groups. The APRs obtained by the template based registration correlated with those using manual method in only mild IPD (r=0.72, P=0.0). The APRs obtained by the template based registration method were significantly different from the normal controls and those with mild or severe IPD (P<0.05), whereas those obtained using the manual method were not significantly different among the three groups (P＞0.1). The reproducibility (rmsCV) of the template based registration method was 7.2% (normal controls：5.2%, mild IPD：4.2%, severe IPD：10.8%), whereas the reproducibility of the manual method was 31% (normal controls：19.7%, mild IPD：21.7%, severe IPD：46.2%). Conclusion：These results show that the use of $^{123}$ I-IPT SPECT for assessing IPD is affected by the methods used to position the striatal ROI. The template based registration method using the standard MR template can be useful in diagnosing IPD and assessing the disease severity with a high reproducibility. Therefore, the template based registration method appears to be a good replacement for the manual method.

• Journal of Korea Multimedia Society
• /
• v.12 no.7
• /
• pp.913-921
• /
• 2009

Quantization is an essential method for compression of video. The quantizer can adjust the bitrate and control the picture quality. Especially, the subjective picture quality can be improved if the concerned region within a video sequence has good picture quality. In this paper, firstly a classification method according to the subjective concerned region within the video sequence is suggested. Also we propose a method that assigns the quantization step-size differentially according to the concerned region within the video. Totally subjective picture quality can be increased by appling the quantization step-size as small value relatively for the concerned region compared with the other regions. We can find the result that the proposed method gives the improved picture quality by assigning differently quantization step-size and the best improvement can be brought when the difference between maximum and minimum values of the quantization step-size in a picture is from 4 to 8.

• Journal of Broadcast Engineering
• /
• v.17 no.4
• /
• pp.670-675
• /
• 2012

If the region of interest (ROI) is set within the picture of image and video and the high quality is provided in ROI compared to Non ROI, then overall subjective picture quality can be increased. ROI extracted by the color camera only increases the calculation complexity and reduces the extraction accuracy. In this paper, we use depth camera to set the ROI and calculate the object distance from camera, then propose a method that the different picture quality is controlled by depending on the distance of an object. That is, we apply a high quantization step size to the far object, but relatively a low quantization step size to the close object, so better picture quality can be provided. Simulation results show that applying the differential quantization step size to the distance of objects by the proposed method can improve the subjective picture quality.