• 대한전자공학회논문지TC
• /
• 제41권9호
• /
• pp.39-45
• /
• 2004

순차적인 관찰값을 바탕으로 하고 신호검출에 소요되는 시간이 고정된 신호검출기의 제작에 관한 방법을 제안하며 이는 하드웨어의 복잡도를 감소시키는 장점이 있다. 제안된 방법은 Voronoi 다이어그램과 Delaunay 분할을 사용한다. 제안된 신호검출기 제작은 또한 고정 지연 트리 검색 검출 (FDTS) 방법에 기반을 둔다. FDTS 는 효율적인 순차적 신호검출 알고리즘이며 심볼간 간섭이 존재하는 채널에서 결정 궤환 등화기법 (DFE)과 결합하여 최적화에 근접한 성능을 보인다. 이러한 접근방법에서는 Voronoi 다이어그램 혹은 등가적으로 Delaunay 분할에 포함된 정보를 활용하여 다차원 유클리드 공간에서의 상대적인 관찰값의 위치를 계산하며 이러한 방법이 효율적인 계산을 유도하는 신호검출기의 제작에 이용된다.

• 한국의학물리학회지:의학물리
• /
• 제16권3호
• /
• pp.111-115
• /
• 2005

본 연구에서는 치료용 광자선의 품질관리를 위하여 다채널 측정기를 개발하였다. 측정기는 여러 개의 전리함들이 삽입된 팬텀으로 구성되어 있다. 각 전리함은 탄소가 도포된 마이크로필름으로 제작된 초소형 평행평판 전리함과 같다. 본 연구에서는 고체 팬텀에 삽입된 6개의 전리함에 대하여 6 MV X-선을 이용하여 전기적 특성을 조사하였다 측정결과 누설전류는 0.5 ph 이내로 안정적이었고, 재현성은 0.5$\%$, 선형성은 0.5$\%$ 이내로 나타났다. 그리고 선량률 효과는 모든 전리함이 0.7$\%$ 이하로 나타났다. 또한 다른 전리함의 영향으로 인한 흡수선량의 변화는 약 0.8$\%$ 이내로 나타났다. 개발된 측정기는 치료용 광자선에 대하여 출력측정 시 선질결정에 활용할 수 있으며 근사적인 깊이선량률의 측정에도 이용될 수 있다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권10호
• /
• pp.4080-4097
• /
• 2020

In recent years, convolutional neural network (CNN) has become the primary method for face detection. But its shortcomings are obvious, such as expensive calculation, heavy model, etc. This makes CNN difficult to use on the mobile devices which have limited computing and storage capabilities. Therefore, the design of lightweight CNN for face detection is becoming more and more important with the popularity of smartphones and mobile Internet. Based on the CPU real-time face detector FaceBoxes, we propose a multi-task lightweight face detector, which has low computing cost and higher detection precision. First, to improve the detection capability, the squeeze and excitation modules are used to extract attention between channels. Then, the textual and semantic information are extracted by shallow networks and deep networks respectively to get rich features. Finally, the landmark detection module is used to improve the detection performance for small faces and provide landmark data for face alignment. Experiments on AFW, FDDB, PASCAL, and WIDER FACE datasets show that our algorithm has achieved significant improvement in the mean average precision. Especially, on the WIDER FACE hard validation set, our algorithm outperforms the mean average precision of FaceBoxes by 7.2%. For VGA-resolution images, the running speed of our algorithm can reach 23FPS on a CPU device.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.77-80
• /
• 1999

This paper deals with a design of a symbol timing recovery circuit of QAM using the interpolation in AWGN channel. To reduce timing jitter and the amount of processing data, we employ MGA (Modified Gardner Algorithm) as a symbol timing error detector which is called NDA(Nondecision Directed Algorithm). We show the characteristics (S-curve and the variance) of timing error detector with the roll-off factor of a shaping filter, which are compared with GA. Also, we compare the BER curve of interpolation method with that of ideal case. The performance of the STR is shown to be close to that of ideal case. This result shows that this method can be useful to implement symbol timing recovery circuit for multi-level modulation.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 하계종합학술대회 논문집(4)
• /
• pp.92-97
• /
• 2000

An AMR(Adaptive Multi-Rate) speech coding algorithm has been adopted as a standard speech codec for IMT-2000. It is based on the algebraic CELP, and consists of eight speech coding modes having the bit rate from 4.75 kbit/s to 12.2 kbit/s. It also contains the VAD(Voice Activity Detector), SCR (Source Controlled Rate) operation, and error concealment scheme for robustness in a radio channel. The bit rate of AMR is changed on a frame basis depending on the channel condition. In this paper, we introduced AMR speech coding algorithm and performed the real-time implementation using TMS320C6201, i.e., a Texas Instrument's fixed-point DSP. With the ANSI C source code released from ETSI and 3GPP, we convert and optimize the program to make it run in real time using the C compiler and assembly language. It is verified that the decoded result of the implemented speech codec on the DSP is identical with the PC simulation result using ANSI C code for test sequences. Also, actual sound input/output test using microphone and speaker demonstrates its proper real-time operation without distortions or delays.

• 한국통신학회논문지
• /
• 제27권3B호
• /
• pp.237-243
• /
• 2002

본 논문에서는 다단계 간섭제거 기법을 적용한 UTRA TDD 모드에서의 채널 추정 기법들의 성능을 다중경로 페이딩 채널 환경에서의 모의 실험을 통해 확인하였다. 또한, UTRA TDD 모드의 장점을 유지할 수 있는 효율적인 Interpolation 기법을 제안하였다. 모의 실험을 수행하여 제안된 기법을 통해 완벽한 채널 추정 조건에서의 다단계 간섭제거 기법의 성능을 얻을 수 있음을 검증하였고. 주어진 BER에서 요구되는 Eb/No를 줄일 수 있음을 확인하였다.

• Nuclear Engineering and Technology
• /
• 제56권5호
• /
• pp.1836-1844
• /
• 2024

The use of digital systems in radiation science has been increased last years in the different knowledge areas, as a detectors, spectrometry, spectroscopy, simulation, etc. This manuscript presents the design and implementation of a low-cost, fully portable multi-channel analyzer for nuclear spectrometry (in situ). The development is based on a 32-bit microcontroller with ARM Cortex-M7, this design is able to digitize and analyze pulses from a radiation detector without the need to transform the input signal with some filter, obtains the maximum height of each of the digitized pulses, segmenting the information into channels to form a histogram and visualizing the LCD screen incorporated in the system. A continuous digitization methodology was used, which is in charge of the DMA and an ADC with a resolution of 12 bits at a speed of 3.6 MSPS. The system has a compact design and can open and save spectra in an SD memory built into the system. The MCA in MCU was tested with a NaI(Tl) Scintillation radiation detector, which allowed us to determine that the spectra obtained are similar compared to commercial MCA's. The results obtained show that the MCA in MCU is efficient for nuclear spectrometry, in addition to being very economical and low power consumption.

• 방사선산업학회지
• /
• 제6권3호
• /
• pp.231-237
• /
• 2012

This paper introduces a gamma-ray measurement system for a transportable tomography which is applicable for an industrial process diagnosis. The gamma-ray measurement system consists of pulse mode operating 72 channel CsI detectors, main AMP-pulse shaper, single channel analyzer, counter and control PC. The CsI crystal is coupled with a PIN diode which is connected to an amplifier and pulse shaper. For a compact design, the amplifier and pulse shaping circuit are included in a single package. 36 sets of CsI detectors are connected to a multi-channel counter through single channel analyzers. A computer controls and collects data from two multi-channel counters. This configuration results in 72 channel counting system in total. The CT rotator and radiation measurement system are controlled by a PC with LabVIEW program. Tomographic data were measured for a phantom by the measurement system and transportable gamma-ray CT. From the experimental data image reconstructions were performed by ML-EM algorithm. The result showed that the CsI detector system can be a suitable component for transportable gamma-ray CT system.

• 한국통신학회논문지
• /
• 제37권3A호
• /
• pp.172-178
• /
• 2012

This paper studies the application of a fuzzy-ARTMAP (FAM) neural network to multi-user detector (MUD) for direct sequence (DS)-code division multiple access (CDMA) system. This method shows new solution for solving the problems, such as complexity and long training, which is found when implementing the previously developed neural-basis MUDs. The proposed FAM based MUD is fast and easy to train and includes capabilities not found in other neural network approaches; a small number of parameters, no requirements for the choice of initial weights, automatic increase of hidden units, no risk of getting trapped in local minima, and the capabilities of adding new data without retraining previously trained data. In simulation studies, binary signals were generated at random in a linear channel with Gaussian noise. The performance of FAM based MUD is compared with other neural net based MUDs in terms of the bit error rate.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1992년도 하계학술대회 논문집 A
• /
• pp.437-440
• /
• 1992

We design a multi-element photo-detector to measure the size of particles using the diffracted light energy distribution. The light energy that is scattered by particles in the collimated laser beam is collected by the Fourier transform lens and directed to the multi-semicircular concentric annular detecters. The scattered profile measured by the photodetector is sampled by a 32 channel analog-to-digital converter. A nonnegative least squares analysis translates the light energy distribution into the corresponding unique particle size distribution.