• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2016년도 춘계학술대회
• /
• pp.363-366
• /
• 2016

본 논문에서는 보행자 검출 시, 교차 상황에서 발생하는 문제 해결을 위한 방법을 제안한다. 영상에서 특정 보행자를 검출하는 동안 다른 보행자와 교차하는 경우, 기존에 검출하던 보행자가 아닌 다른 보행자를 잘못 검출하는 문제가 발생한다. 문제 해결을 위해 제안하는 방법은 다음과 같다. 먼저, 검출할 특정 보행자를 bounding box로 선택하고 해당영역을 템플릿으로 추출한다. HOG를 이용하여 영상에서 보행자들을 검출하고, 후보영역으로 지정한다. 후보영역으로 지정된 보행자들을 앞서 템플릿으로 추출한 특정보행자와 비교하여 검출할 보행자를 최종 선택한다. 비교에는 템플릿 매칭, 히스토그램 비교와 LBP를 이용한다.

• 전자공학회논문지C
• /
• 제34C권10호
• /
• pp.59-66
• /
• 1997

Neural networks has widely been be used for several practical applications such as speech, image processing, and pattern recognition. Thus, a approach to the voltage-controlled current source in areas of neural networks, the key features of CNN in locally connected only to its netighbors. Because the architecture of the interconnection elements between cells in very simple and space invariant, CNNs are suitable for VLSI implementation. In this paper, processing element of digital programmable CNN with variable template memory was implemented using CMOS circuit. CNN PE circuit was designe dto control gain for obtaining the optimal solutions in the CNN output. Performance of operation for 4*4 CNN circuit applied for fixed template and variable template analyzed with the result of simulation using HSPICE tool. As a result of simulations, the proposed variable template method verified to improve performance of operation in comparison with the fixed template method.

• International Journal of Control, Automation, and Systems
• /
• 제6권3호
• /
• pp.413-423
• /
• 2008

The template matching method is used as a simple method to track objects or patterns that we want to search for in the input image data from image sensors. It recognizes a segment with the highest correlation as a target. The concept of this method is similar to that of SNF (Strongest Neighbor Filter) that regards the measurement with the highest signal intensity as target-originated among other measurements. The SNF assumes that the strongest neighbor (SN) measurement in the validation gate originates from the target of interest and the SNF utilizes the SN in the update step of a standard Kalman filter (SKF). The SNF is widely used along with the nearest neighbor filter (NNF), due to computational simplicity in spite of its inconsistency of handling the SN as if it is the true target. Probabilistic Strongest Neighbor Filter for m validated measurements (PSNF-m) accounts for the probability that the SN in the validation gate originates from the target while the SNF assumes at any time that the SN measurement is target-originated. It is known that the PSNF-m is superior to the SNF in performance at a cost of increased computational load. In this paper, we suggest an image tracking algorithm that combines the template matching and the PSNF-m to estimate the states of a tracked target. Computer simulation results are included to demonstrate the performance of the proposed algorithm in comparison with other algorithms.

• 한국정보통신학회논문지
• /
• 제8권4호
• /
• pp.838-845
• /
• 2004

실제 주민등록증을 스캐너로 수집한 후 해당 영상에 대하여 주민등록번호를 인식하였다. 인식을 위한 전처리 과정은 처리 속도를 감안하여 대상부분을 포함 주민등록증의 약 1/8 크기만큼 분할한 후 잡음에 해당하는 홀로그램을 제거하였다. 숫자 인식 방법으로는 원형비교법과 학습법을 병행하였으며 대상 숫자의 단순한 특징 추출을 위해 클러스터링 방식을 사용하였고, 외부 환경에 따라 오인식되는 숫자에 대해 세선화 기법을 병행하여 유사한 숫자간의 유일한 특징으로 구분하였다. 인식에 대한 실험 결과, 숫자 인식에 관한 타 논문의 인식률과 비교하여 양호한 인식률이 도출되었다.

• 천문학회보
• /
• 제41권1호
• /
• pp.73.2-73.2
• /
• 2016

We present the results of KaVA SgrA* observation together with Takahagi(32m), Yamaguchi(32m) and Nobeyama(45m) telescopes at 22 and 43GHz, respectively. In early 2014, G2 cloud was expected to encounter with SgrA* and to make a significant flux variation, but it has not been measured yet. So it's worth to check our amplitude calibration method to confirm if we have a missing flux caused by uncertainty in measuring it. We have tested both a standard method using system noise temperature(Tsys) with antenna gain information, and a template method in order to calibrate antenna gain using nearby maser source. As a result, we found that the latter method is useful for antennas which have inaccurate gain information or poor Tsys measurements, and is especially effective for sources at low elevation like SgrA*. In addition, the comparison shows that the amplitude calibration by standard method can be improved up to 10% with a correction factor using a template method. This result implies we can get more accurate flux from a standard method when any maser source not exists around target.

• 전기전자학회논문지
• /
• 제15권4호
• /
• pp.274-280
• /
• 2011

본 연구에서는 SoC를 이용한 임베디드 시스템에 리눅스 OS 환경을 구축하고 번호판 인식 시스템을 구현하여 그 성능을 측정하였다. 자동차 번호판을 인식하기 위해서는 번호판을 검출하고 검출된 번호판을 보정 한 뒤 각 문자들에 대해 인식을 한다. 번호판 검출 방법으로는 레이블링 기법과 숫자의 특징을 이용하여 검출하였다. 검출된 번호판의 표기되어 있는 숫자들은 각각의 좌표가 있다. 이러한 숫자들의 좌표를 비교하여 영상을 보정하고 템플릿 매칭을 통해 인식을 한다. 그 결과로 번호판의 검출율은 96%, 문자 인식률은 73%, 숫자 인식률은 97%로 나타났다. 인식 시스템은 기존의 PC기반이 아닌 임베디드 보드에서 측정 되었으며 총 인식시간은 약 0.66초가 소요되었다.

• Journal of Electrical Engineering and Technology
• /
• 제9권1호
• /
• pp.190-196
• /
• 2014

In this paper, the characteristics of a defective elliptical aluminum plate are analyzed by three different eddy current testing probes; T/R, T/T, and Impedance probes. The analysis was done by 3D finite element method, and the impedance change of real and imaginary voltage values were analyzed.

• Bulletin of the Korean Chemical Society
• /
• 제34권9호
• /
• pp.2650-2656
• /
• 2013

Mesoporous alumina was synthesized using bayberry tannin (BT) as template. This novel synthesis strategy was based on a precipitation method associated with aluminum nitrate as the source of aluminum in an aqueous system. $N_2$ adsorption/desorption, XRD, SEM and TEM were used to characterize the as-prepared mesoporous alumina. The results showed that the mesoporous alumina possessed crystalline pore wall, high specific surface area, narrow pore distribution and excellent thermal stability. Moreover, the surface area and pore size of the mesoporous alumina can be tuned by changing the experimental parameters. Further, the mesoporous alumina was investigated as the support of palladium catalyst ($Pd-Al_2O_3{^*}$) for the hydrogenation of propenyl, styrene and linoleic acid. For comparison, the reference catalyst ($Pd-Al_2O_3$) prepared without barberry tannin was also employed for the catalytic hydrogenation. The experimental results showed that $Pd-Al_2O_3{^*}$ exhibited the superior catalytic performance than $Pd-Al_2O_3$ for all the investigated substrates, especially for the hydrogenation of linoleic acid with larger molecular.

• Journal of Information Processing Systems
• /
• 제16권3호
• /
• pp.599-611
• /
• 2020

With the application and promotion of biometric technology, biometrics has become more and more important to identity authentication. In order to ensure the privacy of the user, the biometrics cannot be stored or manipulated in plaintext. Aiming at this problem, this paper analyzes and summarizes the scheme and performance of the existing biometric authentication system, and proposes an iris-based ciphertext authentication system based on fully homomorphic encryption using the FV scheme. The implementation of the system is partly powered by Microsoft's SEAL (Simple Encrypted Arithmetic Library). The entire system can complete iris authentication without decrypting the iris feature template, and the database stores the homomorphic ciphertext of the iris feature template. Thus, there is no need to worry about the leakage of the iris feature template. At the same time, the system does not require a trusted center for authentication, and the authentication is completed on the server side directly using the one-time MAC authentication method. Tests have shown that when the system adopts an iris algorithm with a low depth of calculation circuit such as the Hamming distance comparison algorithm, it has good performance, which basically meets the requirements of real application scenarios.

• International Journal of Control, Automation, and Systems
• /
• 제6권6호
• /
• pp.800-808
• /
• 2008

A fingerprint verification system based on a set of invariant moment features and a nonlinear Back Propagation Neural Network(BPNN) verifier is proposed. An image-based method with invariant moment features for fingerprint verification is used to overcome the demerits of traditional minutiae-based methods and other image-based methods. The proposed system contains two stages: an off-line stage for template processing and an on-line stage for testing with input fingerprints. The system preprocesses fingerprints and reliably detects a unique reference point to determine a Region-of-Interest(ROI). A total of four sets of seven invariant moment features are extracted from four partitioned sub-images of an ROI. Matching between the feature vectors of a test fingerprint and those of a template fingerprint in the database is evaluated by a nonlinear BPNN and its performance is compared with other methods in terms of absolute distance as a similarity measure. The experimental results show that the proposed method with BPNN matching has a higher matching accuracy, while the method with absolute distance has a faster matching speed. Comparison results with other famous methods also show that the proposed method outperforms them in verification accuracy.