• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.11
• /
• pp.2555-2562
• /
• 2010

In this paper, we propose the method that extracts sea level measurements graph from the sea level measurements recording paper image with a little interaction. At first, a pixel that is included in the graph area is selected. Then, background pixels are automatically determined using the distance between a selected pixel and other pixels on LAB color space. In each vertical line, a pixel that is the nearest to the selected pixel on LAB color space is extracted and the graph area is determined using that pixels. Experimental results show that the sea level measurements graph can be extracted with a few interaction from the various sea level measurements recording paper images.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.1
• /
• pp.175-182
• /
• 2013

The digitalization of sea level measurement recording which is recorded as analog type is useful for many related oceanology. In this paper, we propose a method which separates the graph area from a sea level measurement recording image. At first, a pixel that is regarded as the pixel which is included in the graph area is selected. Then, many background pixels are separated using the color of the selected pixel. In each vertical line, a pixel is determined as the pixel within the graph area and the graph area is separated from the image using that pixels. Experimental results show that the proposed method in this paper overcome drawbacks of the previous research and can separate the graph area which similar to the graph area of the original image.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.301-304
• /
• 2000

본 논문에서는 디지털 영상에서 얼굴 영상 검출을 위해 픽셀의 퍼지 소속도를 이용하여 신경망으로 학습하는 퍼지 신경망을 이용한 얼굴영상 검출을 제안한다. 입력 영상의 피라미드 영상에서 추출된 20$\times$20 윈도우 템플릿 영상안의 각 픽셀의 소속도로 얼굴 영상 패턴을 학습하여 얼굴 영상을 검출하는 방법은 단순히 영상의 픽셀 값 하나씩만을 고려해서 각 픽셀의 소속도를 고려하여 수행하는 얼굴 영상 분할보다 얼굴 영상을 훨씬 더 정확하고 인식률이 높게 검출해 낼 수 있다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04a
• /
• pp.25-27
• /
• 2002

본 논문에서는 3차원 그래픽의 래스터라이제이션 단계에서 발생하는 메모리 트래픽 문제를 해결하기 위해 사용되는 텍스처 및 픽셀 캐쉬에 대한 성능을 분석하였다. 이를 위해 화면의 해상도, 컬러 정보, 깊이 정보 및 캐쉬 구성의 변화에 따른 이들 캐쉬의 성능변화를 살펴보았으며 실험결과 텍스처 캐쉬와 픽셀 캐쉬의 설계 시에 블록 크기에 의한 영향이 매우 중요함을 알 수 있었다. 특히 픽셀 캐쉬의 경우에는 시간적 지역성은 거의 없으며 매우 큰 공간적 지역성을 보이므로 이를 잘 반영할 수 있는 캐쉬 구조가 필요하다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.1
• /
• pp.39-44
• /
• 2001

본 논문에서는 디지털 영상에서 얼굴 영상 검출을 위해 픽셀의 퍼지 소속도를 이용하여 신경망으로 학습하는 퍼지 신경망을 이용한 얼굴영상 검출을 제안한다. 입력 영상의 피라미드 영상에서 추출된 20$\times$20 윈도우 영상 안의 각 픽셀의 소속도로 얼굴 영상 패턴을 학습하여 얼굴 영상을 검출하는 방법은 단순히 영상의 픽셀 값 하나씩만을 고려해서 각 픽셀의 소속도를 고려하여 수행하는 얼굴 영상 분할보다 얼굴 영상을 더 정확하고 인식률이 높게 검출해 낼 수 있다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.255-256
• /
• 2019

본 논문에서는 초고해상도 기술의 복잡도를 줄이기 위하여 픽셀단위 생략 방법을 제안한다. 제안하는 방법은 픽셀 단위로 수평, 수직 방향의 밝기에 대한 2 차 미분치에 기반하여 생략을 결정한다. 제안하는 방법의 성능 평가를 위하여 가장 간단한 초고해상도 알고리즘인 SRCNN 과 제안하는 방법의 PSNR 비교한다. 그 결과 제안하는 방법이 평균적으로 약 47%의 픽셀이 생략이 되면서 SRCNN 대비 0.2dB PSNR 감소를 보인다.

• Journal of the Korean Society of Radiology
• /
• v.15 no.1
• /
• pp.1-7
• /
• 2021

In order to acquire an image in a positron emission tomography, it is necessary to draw the position coordinates of the scintillation pixels of the detector module measured at the same time. To this end, in a detector module using a plurality of scintillation pixels and a small number of photosensors, it is necessary to obtain a flood image and divide a region of each scintillation pixel to obtain a position of a scintillation pixel interacting with a gamma ray. Alternatively, when the number of scintillation pixels and the number of photosensors to be used are the same, the position coordinates for the position of the scintillation pixels can be directly acquired as digital signal coordinates. A method of using a plurality of scintillation pixels and a small number of photosensors requires a process of obtaining digital signal coordinates requires a plurality of photosensors and a signal processing system. This complicates the signal processing process and raises the cost. To solve this problem, in this study, we developed a method of obtaining digital signal coordinates without performing the process of separating the planar image and region using a plurality of flash pixels and a small number of optical sensors. This is a method of obtaining the position coordinate values of the flash pixels interacting with the gamma ray as a digital signal through a look-up table created through the signals acquired from each flash pixel using the maximum likelihood function. Simulation was performed using DETECT2000, and verification was performed on the proposed method. As a result, accurate digital signal coordinates could be obtained from all the flash pixels, and if this is applied to the existing system, it is considered that faster image acquisition is possible by simplifying the signal processing process.

• 전자공학회논문지 IE
• /
• v.44 no.2
• /
• pp.32-40
• /
• 2007

In this paper, the crease extraction method is proposed to improve the accuracy of feature extraction within the fingerprint image. First of all, for each pixel in fingerprint image, it calculates the average grey level and variance to determine if the current pixel composes the crease, and estimates the direction of crease. Secondly, once the direction of every pixel in crease candidate area is estimated, it is decomposed into 8 different images, depending on their direction. The properties of crease consists of the length of the crease candidate area, the correspondence between the crease direction and the pixel distribution direction, the difference between the ridge direction and the pixel distribution direction, and finally the grey level of the candidate pixels. The proposed method finally extracts the crease from the crease clusters estimated from directional images. In conclusion, applying the proposed method improved the accuracy of overall feature extraction by 91.4% by accurately and precisely extracting the crease from fingerprint image.

• The KIPS Transactions:PartB
• /
• v.11B no.1
• /
• pp.7-12
• /
• 2004

This thesis studies about thinning compensation and the removal of pseudo minutiae using ridge trace. As for the process of thinning compensation, first, it removes unnecessary pixel when 8-neighbor connection sum among the pixels with the crossing count number 6 is not 3. Second, it deals with repeatedly until there is no pixel to remove while tracing the ridge, beginning with the pixel equivalent to the ridge and the pixel with the crossing count number 2 among the 8 pixels around the pixels with the crossing count number 6. When the thinning compensation is finished, it extracts substitute minutiae to use the crossing count number and the 8-neighbor connection sum. Among the extracted substitute minutiaes, it extracts the real minutiae to utilize the pseudo minutiae removal algorithm again. It compares with the existing method for the performance evaluation of proposal method. By the experimental results, The proposal method indicated that a degree of thinning is excellent and a lot of minutiaes were removed.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.11
• /
• pp.95-102
• /
• 2011

Interpolation schemes are used widely in image magnification. Magnified image generated by interpolation scheme is composed of the known pixels in input image and the interpolated pixels estimated from the known pixels in input image. So, as the interpolated pixels are estimated to have locality which exists in real images, the magnified image is much closer to the real image. In this paper, an efficient interpolation scheme was proposed to provide locality for the interpolated pixels by using the characteristics of adjacent pixels in input image. The quality of magnified image using the proposed scheme was improved. In experiment, PSNR(Peak Signal to Noise Ratio) was used to evaluate the performance of the proposed scheme. The PSNR's of the magnified images generated by the proposed scheme were greater than those of the magnified images generated by the previous interpolation methods.