• 스마트미디어저널
• /
• 제8권4호
• /
• pp.64-71
• /
• 2019

The performance of an OMR (Optical Music Recognition) system is usually determined by the characterizing features of the input music score images. Low resolution is one of the main factors leading to degraded image quality. In this paper, we handle the low-resolution problem using the super-resolution technique. We propose the use of a deep neural network with instance normalization to improve the quality of music score images. We apply instance normalization which has proven to be beneficial in single image enhancement. It works better than batch normalization, which shows the effectiveness of shifting the mean and variance of deep features at the instance level. The proposed method provides an end-to-end mapping technique between the high and low-resolution images respectively. New images are then created, in which the resolution is four times higher than the resolution of the original images. Our model has been evaluated with the dataset "DeepScores" and shows that it outperforms other existing methods.

• 대한기계학회논문집
• /
• 제19권6호
• /
• pp.1431-1438
• /
• 1995

In this paper, we present a multiple-view stereo matching method in case of moving in the direction of optical axis with stereo camera. Also we analyze the obtainable depth precision to show that multiple-view stereo increases the virtual baseline with single-view stereo. This method decides candidate points for correspondence in each image pair and then search for the correct combinations of correspondences among them using the geometrical consistency they must satisfy. Adantages of this method are capability in increasing the accuracy in matching by using the multiple stereo images and less computation due to local processing. This method computes 3-D depth by averaging the depth obtained in each multiple-view stereo. We show that the resulting depth has more precision than depth obtainable by each independent stereo when the position of image feature is uncertain due to image noise. This paper first defines a multipleview stereo agorithm in case of moving in the direction of optical axis with stereo camera and analyze the obtainable precision of computed depth. Then we represent the effect of removing the incorrect matching candidate and precision enhancement with experimental result.

• 한국연소학회지
• /
• 제1권1호
• /
• pp.65-73
• /
• 1996

Flame propagation in a four-valve spark-ignition optical engine was visualized under lean-bum conditions with A/F=18 at 2000rpm. The early flame development in a four-valve pentroof-chamber single-cylinder engine was examined with imaging of the laser-induced Mie scattered light using an image-intensified CCD camera. Flame profiles along the line-of-sight were also visualized through a quartz piston window. Two-dimensional flame structures were visualized with a Proxitronic HF-1 fast motion camera system by Mie scattering from titanium dioxide particles along a planar laser sheet generated by a copper vapor laser. The flame propagation images were subsequently analysed with an image processing programme to obtain information about the flame structure under different tumble flow conditions generated by sleeved and non-sleeved intake ports. This allowed enhancement of the flame images and calculation of the enflamed area, and the displacement of its center, as a function of the tumble flow induced by the pentroof-chamber in the vicinity of spark plug. Image processing of the early flame development quantified the correlation between flame and flow characteristics near the spark plug at the time of ignition which has been known to be one of the most important factors in cyclic combustion variations in lean-burn engines. The results were also compared with direct flame images obtained from the natural flame luminosity of the lean mixture.

• ETRI Journal
• /
• 제39권5호
• /
• pp.683-694
• /
• 2017

Several studies have focused on multimedia transmission over wireless sensor networks (WSNs). In this paper, we propose a comprehensive and robust model to transmit images over cognitive radio WSNs (CRWSNs). We estimate the spectrum sensing frequency and evaluate its impact on the peak signal-to-noise ratio (PSNR). To enhance the PSNR, we attempt to maximize the number of pixels delivered to the receiver. To increase the probability of successful image transmission within the maximum allowed time, we minimize the average number of packets remaining in the send buffer. We use both single- and multi-channel transmissions by focusing on critical transmission events, namely hand-off (HO), No-HO, and timeout events. We deploy our advanced updating method, the dynamic parameter updating procedure, to guarantee the dynamic adaptation of model parameters to the events. In addition, we introduce our ranking method, named minimum remaining packet best channel selection, to enable us to rank and select the best channel to improve the system performance. Finally, we show the capability of our proposed image scrambling and filtering approach to achieve noticeable PSNR improvement.

• 국제학술발표논문집
• /
• The 10th International Conference on Construction Engineering and Project Management
• /
• pp.1065-1072
• /
• 2024

Previous approaches to 3D excavator pose estimation via synthetic data training utilized a single virtual excavator model, low polygon objects, relatively poor textures, and few background objects, which led to reduced accuracy when the resulting models were tested on differing excavator types and more complex backgrounds. To address these limitations, the authors present a realism-centric synthetization and labeling approach that synthesizes results with improved image quality, more detailed excavator models, additional excavator types, and complex background conditions. Additionally, the data generated includes dense pose labels and depth maps for the excavator models. Utilizing the realism-centric generation method, the authors achieved significantly greater image detail, excavator variety, and background complexity for potentially improved labeling accuracy. The dense pose labels, featuring fifty points instead of the conventional four to six, could allow inferences to be made from unclear excavator pose estimates. The synthesized depth maps could be utilized in a variety of DNN applications, including multi-modal data integration and object detection. Our next step involves training and testing DNN models that would quantify the degree of accuracy enhancement achieved by increased image quality, excavator diversity, and background complexity, helping lay the groundwork for broader application of synthetic models in construction robotics and automated project management.

• Journal of Electrical Engineering and Technology
• /
• 제2권2호
• /
• pp.274-279
• /
• 2007

This study focuses on the enhancement of maskless photolithography as well as the peptide synthesis application with single crystalline silicon micromirrors. A single crystalline silicon micromirror array has been designed and fabricated in order to improve its application to the peptide synthesis. A micromirror rotates about ${\pm}\;9^{\circ}$ at the pull-in voltage, which can range from 90.7 V to 115.1 V. A $210\;{\mu}m-by-210\;{\mu}m$ micromirror device with $270\;{\mu}m$ mirror pitch meets the requirements of an adequately precise separation for peptide synthesis. Synthetic 16 by 16 peptide array corresponds to the same number of micromirrors. The large size of peptide pattern and the separation facilitate biochip experiments using fluorescence assay. The peptide pattern has been synthesized on the GPTS-PEG200 surface with BSA-blocking and thereupon the background was acetylated to reject non-specific bindings. Hence, an averaged slope at the pattern edge has been distinguishably improved in comparison to patterning results from an aluminum micromirror.

• 대한방사선기술학회지:방사선기술과학
• /
• 제44권5호
• /
• pp.465-471
• /
• 2021

Dual-energy X-ray imaging (DEI) techniques can provide X-ray images that a certain material is suppressed or emphasized by combining two X-ray images obtained from two different x-ray spectrum. In this paper, a single-shot DEI, which uses stacked two detectors (i.e., multilayer detector), is proposed to reduce the patient dose and increase throughput in angiography. The polymethyl methacrylate (PMMA) and aluminum (Al) were selected as two basis materials for material decomposition, and material-specific images are reconstructed as a vector combination of these two materials. We investigate the contrast and noise performance of material-decomposed images using iodine phantoms with various concentrations and diameters. The single-shot DEI shows comparable performances to the conventional dual-shot DEI. In particular, the single-shot DEI shows edge enhancement in material-decomposed images due to the different spatial-resolution characteristics of upper and lower detectors. This study could be useful for designing the multilayer detector including scintillators and energy-separation filter for angiography purposes.

• 한국전자통신학회논문지
• /
• 제14권1호
• /
• pp.257-264
• /
• 2019

본 논문에서는 픽셀 기반 joint BDCP (bright and dark channel prior)와 계층적 양방향 필터를 적용하여 저 복잡도를 갖는 단일 영상 기반 안개 제거 기법을 제안한다. 픽셀 기반 joint BDCP는 기존의 패치 기반 DCP에 비해 연산량을 감소시키고, 픽셀 단위의 안개값 예측을 가능하게 하여 전달량 추정의 정확성을 높인다. 또한 에지를 보존하면서 평탄화 성능이 우수한 양방향 필터를 사용하여 전달량을 정련함으로써 후광 효과(halo effect)를 줄이고, 에지 성분에 대한 계층적 적용을 통해 반복 적용에 의한 연산량의 증가를 방지한다. 안개 성분이 포함된 다양한 영상에 대해 수행한 실험 결과는 제안하는 기법이 기존의 기법에 비해 우수한 안개 제거 성능을 보이면서 저 복잡도로 실행되어 다양한 분야에 응용될 수 있음을 나타낸다.

• 디지털정책학회지
• /
• 제2권3호
• /
• pp.1-7
• /
• 2023

본 논문은 단일 영상을 이용하여 초해상도 방법을 수행하기 위해 질감-공간 영역을 분리한 뒤 세부정보를 중심으로 특징을 분류하는 방법을 제안한다. CNN(Convolutional Neural Network) 기반의 초해상도는 세부정보를 개선하기 위한 특징 추정 과정에서의 복잡한 절차와 중복된 특징 정보의 생성으로 인해 초해상도에서 가장 중요한 기준인 품질 저하가 발생할 수 있다. 제안하는 방법은 절차적 복잡성을 줄이고 중복 특징 정보의 생성을 최소화하여 초해상도 결과의 품질을 개선하기 위해 입력 영상을 질감과 공간의 두 채널로 분리하였다. 질감 채널에서는 세부정보 복원을 위해 다중스케일로 변환한 영상에 단계별 skip-connection을 적용한 잔차 블록 구조를 적용하여 특징 정제 과정을 수행함으로써 특징 추출을 개선하였고, 공간 채널에서는 평활화된 형태의 특징을 활용하여 잡음을 제거하고 구조적 특징을 유지하도록 하였다. 제안하는 방법을 이용해 실험한 결과 기존 초해상도 방법대비 PSNR 및 SSIM 성능 평가에서 향상된 결과를 보여 품질이 개선됨을 확인할 수 있었다.

• 전기전자학회논문지
• /
• 제20권1호
• /
• pp.89-93
• /
• 2016

본 논문에서는 다양한 영상 전처리 알고리즘들을 적용할 수 있고, 영상 인식과 같이 ISP 응용 프로그램을 병렬로 처리 가능한 SIMT(Single Instruction Multiple Threads) 기반 Image Signal Processor를 제안하였다. 기존의 ISP는 입력 영상의 품질 개선을 위하여 영상 개선 알고리즘이 하드웨어로 설계되어 처리 속도는 빠르지만 다양한 영상 처리 알고리즘에 따라 성능 최적화에 어려움이 있었다. 제안한 ISP는 명령어를 기반으로 한 프로세서로서 다양한 영상 처리 알고리즘을 수행하고 SIMT 구조를 적용하여 알고리즘을 병렬로 수행해 성능을 개선하였다. 제안하는 ISP를 검증하기 위해 Xilinx Virtex-7을 탑재한 VC707 Board를 사용하였으며 cell multicore processor와 비교했을 경우 수행시간이 약 71%, ARM Cortex-A9과 ARM Cortex-A15와 비교하였을 경우 각각 63%, 33% 성능을 개선하였다.