• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.3
• /
• pp.240-248
• /
• 2013

In this paper, we propose a super-resolution algorithm using an adaptive weighted interpolation(AWI) and discrete wavelet transform(DWT). In general, super-resolution algorithms for single-image, probability based operations have been used for searching high-frequency components. Consequently, the complexity of the algorithm is increased and it causes the increase of processing time. In the proposed algorithm, we first find high-frequency sub-bands by using DWT. Then we apply an AWI to the obtained high-frequency sub-bands to make them have the same size as the input image. Now, the interpolated high-frequency sub-bands and input image are properly combined and perform the inverse DWT. For the experiments, we use the down-sampled version of the original image($512{\times}512$) as a test image($256{\times}256$). Through experiment, we confirm the improved efficiency of the proposed algorithm comparing with interpolation algorithms and also save the processing time comparing with the probability based algorithms even with the similar performance.

• The Journal of the Korea Contents Association
• /
• v.7 no.11
• /
• pp.34-42
• /
• 2007

In this paper, we propose a staff-line detection and removal algorithm from a music score image obtained by a mobile phone camera. As a preprocessing technique to recognize a music score image, staff-line detection and removal should be efficiently applied to the skewed or curved images. The proposed method detects a staff-line by dividing a staff according to the degree of distortion. The number of division is calculated by dividing a staff repletely until an average of differences of y coordinates in every divided position is smaller than a threshold. Therefore, the number of division can be adaptively estimated according to the degree of the distortion. For an experiment, we make various kinds of images by rotating one from $1^{\circ}\;to\;3^{\circ}$ or curving slightly upward. The results show that the proposed method performed well on the experiment images.

• Smart Media Journal
• /
• v.5 no.1
• /
• pp.24-29
• /
• 2016

To satisfy the needs of high-level intelligent surveillance system, it shall be able to extract objects and classify to identify precise information on the object. The representative method to identify one's identity is face recognition that is caused a change in the recognition rate according to environmental factors such as illumination, background and angle of camera. In this paper, we analyze the robust face recognition of face image by changing the distance through a variety of experiments. The experiment was conducted by real face images of 1m to 5m. The method of face recognition based on Linear Discriminant Analysis show the best performance in average 75.4% when a large number of face images per one person is used for training. However, face recognition based on Convolution Neural Network show the best performance in average 69.8% when the number of face images per one person is less than five. In addition, rate of low resolution face recognition decrease rapidly when the size of the face image is smaller than $15{\times}15$.

• Journal of the Korea Convergence Society
• /
• v.12 no.1
• /
• pp.29-35
• /
• 2021

License plate recognition plays a key role in intelligent transportation systems. Therefore, it is a very important process to efficiently detect the number and character areas. In this paper, we propose a method to effectively detect license plate number area by applying deep learning and semantic image segmentation algorithm. The proposed method is an algorithm that detects number and text areas directly from the license plate without preprocessing such as pixel projection. The license plate image was acquired from a fixed camera installed on the road, and was used in various real situations taking into account both weather and lighting changes. The input images was normalized to reduce the color change, and the deep learning neural networks used in the experiment were Vgg16, Vgg19, ResNet18, and ResNet50. To examine the performance of the proposed method, we experimented with 500 license plate images. 300 sheets were used for learning and 200 sheets were used for testing. As a result of computer simulation, it was the best when using ResNet50, and 95.77% accuracy was obtained.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2010.07a
• /
• pp.250-251
• /
• 2010

디지털 카메라의 보급으로 이미지처리 분야에서 정지영상을 이용한 다양한 기술 개발이 화두가 되고 있다. 스테레오 영상은 정지영상보다 소비자의 시각적 욕구를 충족시킬 수 있는 영상을 표현하기 때문에 스테레오 영상기술에 대한 관심이 높아지고 있다. 본 논문에서는 하나의 카메라로 같은 객체를 다른 위치에서 찍은 2장의 정지영상을 통해 스테레오 영상을 제작하는 방법을 제안한다. 실험 영상으로 디지털카메라로 찍은 좌측 영상과 우측영상을 사용한다. 두 영상의 제어점이 될 코너를 검출한 후, 유클리드의 좌표로 바꿔준다. 이 좌표들을 통해 각 제어점에 인접해 있는 좌표 4개를 추출한다. 이 인접 좌표들이 우측 정지 영상의 인접 좌표에 매칭 되는 횟수를 계산하여, 가장 많은 매칭 좌표를 갖는 스케일 요소로 좌측 정지영상을 회전과 이동시켜 목적 영상인 우측 영상에 매칭시킴으로써 스테레오 영상을 구현한다.

• Science of Emotion and Sensibility
• /
• v.23 no.3
• /
• pp.47-62
• /
• 2020

The snake detection theory posits that, due to competition with snakes, the primate visual system has been evolved to detect camouflaged snakes. Specifically, one of its hypotheses states that the subcortical visual pathway mainly consisting of koniocellular cells enables humans to automatically detect the threat of snakes without consuming mental resources. Here we tested the hypothesis by comparing human participants' responses to snakes with those to fearful faces and flowers. Participants viewed either original images or converted ones, which lacked the differences in color, luminance, contrast, and spatial frequency energies between categories. While participants in Experiment 1 produced valence and arousal ratings to each image, those in Experiment 2 detected target images in the breaking continuous flash suppression (bCFS) paradigm. As a result, visual factors influenced the responses to snakes most strongly. After minimizing visual differences, snakes were rated as being less negative and less arousing, and detected more slowly from suppression. In contrast, the images of the other categories were less affected by image conversion. In particular, fearful faces were rated as greater threats and detected more quickly than other categories. In addition, for snakes, changes in arousal ratings and those in bCFS response times were negatively correlated: Those snake images, the arousal ratings of which decreased, produced increased detection latency. These findings suggest that the influence of snakes on human responses to threat is limited relative to fearful faces, and that detection responses in bCFS share common processing mechanisms with conscious ratings. In conclusion, the current study calls into question the assumption that snake detection in humans is a product of unconscious subcortical visual processing.

• Korean Journal of Remote Sensing
• /
• v.36 no.5_1
• /
• pp.807-821
• /
• 2020

The objective of this study is to evaluate the applicability of representative spatio-temporal fusion models developed for the fusion of mid- and low-resolution satellite images in order to construct a set of time-series high-resolution images for crop monitoring. Particularly, the effects of the characteristics of input image pairs on the prediction performance are investigated by considering the principle of spatio-temporal fusion. An experiment on the fusion of multi-temporal Sentinel-2 and RapidEye images in agricultural fields was conducted to evaluate the prediction performance. Three representative fusion models, including Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM), SParse-representation-based SpatioTemporal reflectance Fusion Model (SPSTFM), and Flexible Spatiotemporal DAta Fusion (FSDAF), were applied to this comparative experiment. The three spatio-temporal fusion models exhibited different prediction performance in terms of prediction errors and spatial similarity. However, regardless of the model types, the correlation between coarse resolution images acquired on the pair dates and the prediction date was more significant than the difference between the pair dates and the prediction date to improve the prediction performance. In addition, using vegetation index as input for spatio-temporal fusion showed better prediction performance by alleviating error propagation problems, compared with using fused reflectance values in the calculation of vegetation index. These experimental results can be used as basic information for both the selection of optimal image pairs and input types, and the development of an advanced model in spatio-temporal fusion for crop monitoring.

• Proceedings of the IEEK Conference
• /
• 2001.06d
• /
• pp.235-238
• /
• 2001

최근 효과적인 내용기반 영상검색을 위해 특징 추출 방법이 많이 연구되고 있다. 특히 칼라 정보를 이용하여 특징을 얻는 방법은 여러 가지 장점 때문에 많이 사용되고 있다 본 논문에서는 칼라 코렐로그램(color correlogram) 기반의 새로운 특징 추출 방법을 제안한다. 제안한 방법은 웨이브릿 변환 계수를 사용하여 영상을 복잡한 영역과 그렇지 않은 영역으로 분할하고, 각 영역의 칼라 코렐로그램을 영상의 특징으로 사용해 영상을 검색하는 방법이다. 제안한 방법으로 영상을 검색하는 방법은 기존의 칼라 코렐로그램을 이용한 방법보다 성능이 우수함을 실험에서 확인할 수 있었다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2000.11b
• /
• pp.87-92
• /
• 2000

본 논문에서는 컨텍스트 기반 계층적 열거 부호화를 이용한 무손실 이진 영상 압축 알고리즘을 제안한다. 이진 영상내에 존재하는 인접한 화소간의 상호상관성을 이용하여 이진 영상을 1차원의 수열로 재구성하고, 이에 대해 계층적 열거 부호화를 실행한다. 제안하는 알고리즘은 덧셈 및 비교 연산만으로 구현이 가능하므로 그 복잡도가 매우 낮을 뿐만 아니라, CCITT 테스트 영상을 대상으로 한 부호화 성능 실험에서 우수한 성능을 나타낸다. 부호화 성능 비교에서 이진 영상 부호화 국제표준인 JBIG, G3, G4 및 GIF에 비해 우수한 압축 성능을 보인다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 1998.04a
• /
• pp.148-153
• /
• 1998

칼라 역 해프토닝(Color inverse halftoning)은 해프토닝 칼라 영상을 시각적으로 보다 자연스러운 연속계조 칼라영상으로 변환해 주는 방법이다. 본 논문에서는, 분리된 각 칼라 채널 영상에서 나타나는 해프톤 셀 패턴을 효과적으로 제거할 수 있는 새로운 평활화 마스크를 제안하고, 이를 칼라역 해프토닝에 활용하였다. 제안한 평활화 마스크는 기존의 평활화 마스크가 잘 제거하지 못했던 해프톤, 셀 패턴을 시각적으로 보다 자연스럽게 평활해 줄 수 있으며, 마스크의 특성을 평활화하고자 하는 채널 영상에 적합하게 조정할 수 있도록 설계하였다. 실험을 통해, 평활화된 채널 영상과 그 스펙트럼을 비교 분석함으로써 제안한 방법의 유용성을 확인하였다. 이 방법은 전자 출판, 칼라 팩스, 해프톤 영상의 압축 등의 분야에 활용될 수 있다.