• The Journal of the Korea Contents Association
• /
• v.21 no.8
• /
• pp.1-9
• /
• 2021

Deep learning in computer vision has made accelerated improvement over a short period but large-scale learning data and computing power are still essential that required time-consuming trial and error tasks are involved to derive an optimal network model. In this study, we propose a similar image classification performance improvement method based on CR (Confusion Rate) that considers only the characteristics of the data itself regardless of network optimization or data reinforcement. The proposed method is a technique that improves the performance of the deep learning model by calculating the CRs for images in a dataset with similar characteristics and reflecting it in the weight of the Loss Function. Also, the CR-based recognition method is advantageous for image identification with high similarity because it enables image recognition in consideration of similarity between classes. As a result of applying the proposed method to the Resnet18 model, it showed a performance improvement of 0.22% in HanDB and 3.38% in Animal-10N. The proposed method is expected to be the basis for artificial intelligence research using noisy labeled data accompanying large-scale learning data.

• Proceedings of the IEEK Conference
• /
• 2003.11a
• /
• pp.385-388
• /
• 2003

In this paper, we propose a new saturation enhancement algorithm which is processed on the new color space, called Normalized YCbCr(NYCbCr). The algorithm consists of two processing unit. One is color space conversion from YCbCr to NYCbCr, and the other is using adaptive saturation mapping function(ASMF). NYCbCr color space is designed to prevent shortcomings such as luminance and hue shift of YCbCr color space and by saturation enhancement. ASMF is effective to enhance saturation properly for each image and to protect low saturation regions of color images from over-saturation. we verified our method using several color images. Experimental results show that the proposed method enhance the saturation with minimizing Luminance and Hue shift.

• Korean Journal of Digital Imaging in Medicine
• /
• v.2 no.1
• /
• pp.92-95
• /
• 1996

Computed Radiography(CR) is a relatively new technology that relies on an image plate(IP) as an alternate x-ray sensor to screen/film. Standard CR cassettes do not have lead foil behind the IP to control scatter radiation. The result of this study indicate that such control is needed. In most screen/film cassettes, that lines the rear of the cassette eliminates back scatter radiation. This study was performed to Investigate on the effects of back scatter in CR images by size of exposure field, distance between the CR cassette and the wall of radiography room. 1. It showed artifacts from hinges and clips located on the back of CR cassette by back scatter radiation. 2. The greater effects of back scatter radiation in CR images was attributed to the greater size of exposure field and the longer distance between the CR cassette and the wall of radiography room.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1528-1535
• /
• 1997

Many radiological modalities has been applied to medicine as a basic fundamental diagnosis and therapy recently. The prevalence of computer systems affect most images to be digitized. However conventional X-ray film images are not digital images eventhough they covers 70% of all radiologica images. This is the hinderacne of building PACS. In this paper all radiological digital imaging parts such as DSA. CR. MRI. SPECT. PET and ultrasonography were briefly introduced and the applications were described. In brief digital radiography contribute to enhance the medical service quality. And the digital substituition of conventional X-ray film image is inevitable.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2008.10a
• /
• pp.700-702
• /
• 2008

A new evaluation method which named 'Security Viewing Angle Factor (SVAF)' was suggested for VIC (Viewing-angle Image Control) technology to quantify a visual hiding performance of LCD. Currently, Contrast Ration (CR) is the general method to evaluate VIC technology. However, CR shows a significant difference when compared with the result of human visual system. Furthermore, a perceptional evaluation method and its conditions were established.

• Journal of Radiation Protection and Research
• /
• v.36 no.4
• /
• pp.216-222
• /
• 2011

Lab effectively supplemented the effects of outside radiation on image plates in the process of image acquisition of CR (computed radiography) systems and conducted for effective utilization in the case of clinical application. For this, Lab classified the storage places and time periods of image plates and compared and analyzed the differences between small dark spots. Lab also assessed the concentration distribution within the boundaries of images. Lab compared and measured the number of dark spots in a light room and a dark room depending on the storage places of image plates and found that dark spots slightly increased in an image plate when stored in a light room on the first and second days. Dark spots increased in proportion to the length of time stored. In the case of the image plate stored in a dark room, the number of dark spots remarkably decreased. With regard to picture quality as related to the location of image plates, the damage to picture quality could be reduced by locating regions of interest in the center. With regard to differences in sharpness following changes in the thickness of subjects, fewer scatter rays occurred and sharpness improved by reducing the thickness of subjects as much as possible. To get medical images of excellent quality, image plates should be managed effectively and it is desirable to keep images plates in dark iron plate boxes and not to expose them to outside radiation for a long time.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.2
• /
• pp.120-126
• /
• 2016

In this study, we propose a method for effectively detecting and recognizing the face in image using RBFNNs pattern classifier and HCbCr-based skin color feature. Skin color detection is computationally rapid and is robust to pattern variation for face detection, however, the objects with similar colors can be mistakenly detected as face. Thus, in order to enhance the accuracy of the skin detection, we take into consideration the combination of the H and CbCr components jointly obtained from both HSI and YCbCr color space. Then, the exact location of the face is found from the candidate region of skin color by detecting the eyes through the Haar-like feature. Finally, the face recognition is performed by using the proposed FCM-based RBFNNs pattern classifier. We show the results as well as computer simulation experiments carried out by using the image database of Cambridge ICPR.

• The Journal of Art Theory & Practice
• /
• no.16
• /
• pp.225-256
• /
• 2013

As the great catastrophe in the modern and contemporary history of China, the Cultural Revolution(CR) is an object, which must have sutured the past of darkness. At the same time it is a continuous event and also a scar of memory. In other words, for history is "a dialog between the past and the reality"(E.H.Kar), CR intervenes in the reality and, on the contrary, the reality recomposes CR. From this point of view, CR is a historical event, which so far is not ended, and it is an object of memory, which is still being composed at the moment. As the saying: "Poetry is greater than history"(Aristoteles), artistic works more intensively reflect the past. The works related to CR can not be an exception. And CR is endlessly exposed in the contemporary Chinese fine arts and the works of the contemporary Chinese artists-Wang Guangyi, Yue Minjun, Zhang Xiaogang and others are proved to be those who suffer from the trauma of CR and who feel no liberty from CR. For example, CR probably is a symbol showing the "identity" of the Chinese artists. And the diversity of the symbol is the experience and pattern of the dialog between artistic works and CR (i.e., intervention in reality). For example, with withering of grand-narrative and appearance of micro-narrative, the CR critical works of Guan Zhoudao were the root of the Chinese fine arts in the late 70s and early 80s. In the contemporary cultural situation, the literary works about CR actively analyzed the history (CR) from the personal point of views and explained in the way of monolog and micro-method led the 1990s' works. In this way they tried to recompose the history "randomly", like looking at reality with their own eyes. In this process, CR is continuously exposing new features and the real facts are appearing before us as unfamiliar phenomena. This is a way of combination and "reappearance" of contemporary arts and reality. In conclusion, the purpose of this article is to make it possible to see a section of the contemporary Chinese fine arts through the study of the icon image of CR and to analyze the way of fine arts recomposing the history and the intervening in the reality. In this sense, the author has entitled the article "Icon and Form". It means how to reshape (the present) the typically formed icon of the CR (the past).

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.04a
• /
• pp.1042-1046
• /
• 2000

본 논문에서는 질감 정보를 나타낼 수 있는 Block-based 에지 히스토그램과 색상 정보를 표현할 수 있는 DCT 를 이용한 칼라 영상 검색 방법을 제안한다. 제안된 방법은 최소의 특징량으로 최대의 검색효율을 얻기 위해 YCbCr 칼라 모델상에서 Y 영상으로부터는 전체적인 영상에 대한 히스토그램과 에지 히스토그램을 특징량으로 추출하고 Cb, Cr 영상으로부터는 DCT 계수를 특징량으로 추출하여 칼라 영상을 검색한다. 이는 칼라와 질감을 동시에 고려하면서 특징량의 크기가 적어 웹, 대용량 검색 시스템 및 동영상 검색에 적합하다. 성능 평가는 MPEG-7 의 칼라 특징자들의 성능평가를 위해 사용된 S1 및 S3 그룹 영상을 대상으로 실험하였으며 제안한 복합 특징량은 칼라 영상 검색에서 우수한 성능을 나타냄을 실험으로 확인 하였다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.5
• /
• pp.103-110
• /
• 2008

A reliable extraction of welded area is the precedent task before the detection of weld defects in industrial radiography. This paper describes an attempt to detect and extract the welded features of steel tubes from the computed radiography(CR) images. The statistical properties are first analyzed on over 160 sample radiographic images which represent either weld or non-weld area to identify the differences between them. The analysis is then proceeded by pattern classification to determine the clustering parameters. These parameters are the width, the functional match, and continuity. The observed weld image is processed line by line to calculate these parameters for each flexible moving window in line image pixel set. The local statistic line-clustering method is used as the classifier to recognize each window data as weld or non-weld cluster. The sequential procedure is to track the edge lines between two distinct regions by iterative calculation of threshold, and it results in extracting the weld feature. Our methodology is concluded to be effective after experiment with CR weld images.