• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.06a
• /
• pp.118-121
• /
• 1998

In this paper, we propose another paradigm(QNBP) to be capable of overcoming Limitations of the traditional backpropagation(SDBP). QNBPis based on the method of Quasi -Newton(variable metric) with the nomalized direction vectors and computes step size through the linear search. Simulation results showed that QNBP was definitely superior to both the stochasitc SDBP and the deterministic SDBP in terms of accuracy and rate of convergence and might sumount the problem of local minima. and there was no different between DFP+SR1 and BFGS+SR1 combined algrothms in QNBP.

• Journal of Communications and Networks
• /
• v.18 no.5
• /
• pp.735-743
• /
• 2016

This paper proposes a direction-of-arrival (DOA)-based beamforming approach for multi-cell massive multiple-input multiple-output systems with uniform rectangular arrays (URAs). The proposed approach utilizes the steering vectors of the URA to form a basis of the spatial space and selects the partial space for beamforming according to the DOA information. As a result, the proposed approach is of lower computational complexity than the existing methods which utilize the channel covariance matrices. Moreover, the analysis demonstrates that the proposed approach can eliminate the interference in the limit of infinite number of the URA antennas. Since the proposed approach utilizes the multipaths to enhance the signal rather than discarding them, the proposed approach is of better performance than the existing low-complexity method, which is verified by the simulation results.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.8B
• /
• pp.1500-1505
• /
• 1999

In this paper, we presented the stereo algorithm for 3-D object recognition. In order to solve the problem for matching time in existed methods, we proposed the method which used the moving direction vector. On the other hand, after we extracted the moving vectors by moving direction of objects, rotated object was matched on axis of it. Using the Hough transform, we obtained the 2-D synthesed image as reference images corresponding to the rate of moving, and then compared with the unknown input images.

• Journal of Information Processing Systems
• /
• v.16 no.4
• /
• pp.774-783
• /
• 2020

For the issue of subsurface target localization by reverse projection, a new approach of target localization with different distances based on symmetric sub-array multiple-input multiple-output (MIMO) radar is proposed in this paper. By utilizing the particularity of structure of the two symmetric sub-arrays, the received signals are jointly reconstructed to eliminate the distance information from the steering vectors. The distance-independent direction of arrival (DOA) estimates are acquired, and the localizations of subsurface targets with different distances are realized by reverse projection. According to the localization mechanism and application characteristics of the proposed algorithm, the grid zooming method based on spatial segmentation is used to optimize the locaiton efficiency. Simulation results demonstrate the effectiveness of the proposed localization method and optimization scheme.

• Journal of Korea Multimedia Society
• /
• v.22 no.5
• /
• pp.580-587
• /
• 2019

In this paper, a plane detection method using depth information is proposed. 3-D characteristics of a pixel are defined as a direction and length of a normal vector whose is calculated from a plane consisting of a local region centered on the pixel. Image coordinates of each pixel are transformed to 3-D coordinates in order to obtain the local planes. Regions of each plane are detected by calculating similarity of the 3-D characteristics. The similarity of the characteristics consists of direction and distance similarities of normal vectors. If the similarity of the characteristics between two adjacent pixels is enough high, the two pixels are regarded as consisting of same plane. Simulation results show that the proposed method using the depth picture is more accurate for detecting plane areas than the conventional method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1867-1870
• /
• 2005

The celestial navigation is one of alternatives to GPS system and can be used as a backup of GPS. In the celestial navigation system using more than two star trackers, the vehicle's ground position can be solved based on the star trackers' attitude information if the vehicle's local vertical or horizontal angle is given. In order to determine accurate ground position of flight vehicle, the high accurate local vertical angle measurement is one of the most important factors for navigation performance. In this paper, the Earth geophysical deflection was analyzed in the assumption of using the modern electrolyte tilt sensor as a local vertical sensor for celestial navigation system. According to the tilt sensor principle, the sensor measures the tilt angle from gravity direction which depends on the Earth geoid surface at a given position. In order to determine the local vertical angle from tilt sensor measurement, the relationship between the direction of gravity and the direction of the Earth center should be analyzed. Using a precision orbit determination software which includes the JGM-3 Earth geoid model, the direction of the Earth center and the direction of gravity are extracted and analyzed. Appling vector inner product and cross product to the both extracted vectors, the magnitude and phase of deflection angle between the direction of gravity and the direction of the Earth center are achieved successfully. And the result shows that the angle differences vary as a function of latitude and altitude. The maximum 0.094$^{circ}$angle difference occurs at 45$^{circ}$latitude in case of 1000 Km altitude condition.

• Journal of Broadcast Engineering
• /
• v.6 no.1
• /
• pp.50-57
• /
• 2001

A digital watermarking is a newly developed scheme to embed invisible or inaudible information Into the host data in order to insist the copyright of the owner or the creator. This paper describes a robust data embedding scheme that employs inner product and adaptive quantization. Compared to the previous works for digital watermarking, our proposed scheme can embed relatively large amount of Information, since a secrete key Is not directly relaxed to the watermark data. A secret key is used for the design of random direction vectors. which are taken Inner product with the DCT transformed feature set data. In odder to achieve robustness against malicious attacks. we exploit the Properties of human visual system In designing the random direction vectors which behave as embedded noises. Experimental results show that we can recover the embedded information without utilizing the original host data. We also demonstrate that the ownership assertion is possible even though The watermarked data may undergo common signal processing operations, such as JPEG compression. clopping. and filtering.

• Journal of Advanced Navigation Technology
• /
• v.20 no.5
• /
• pp.433-439
• /
• 2016

Several low-cost global navigation satellite system (GNSS) receivers do not support general range-domain correction, and DGNSS-CP (differential GNSS) method had been suggested to solve this problem. It improves its position accuracy by projecting range-domain corrections to the position-domain and then differentiating the stand-alone position by the projected correction. To project the range-domain correction, line-of-sight vectors from the receiver to each satellite should be calculated. The line-of-sight vectors can be obtained from GNSS broadcast ephemeris data or satellite direction information, and this paper shows positioning performance for the two methods. Stand-alone positioning result provided from Septentrio PolaRx4 Pro receiver was used to show the difference. The satellite direction information can reduce the computing load for the DGNSS-CP by 1/15, even though its root mean square(RMS) of position error is bigger than that of ephemeris data by 0.1m.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.39 no.1
• /
• pp.60-66
• /
• 2002

In this paper, an efficient content-based rotation-invariant retrieval of the trademarks is proposed using the edge-direction histogram for a principal symmetry axis and the moment invariants. Rotation invariant retrieval of trademarks is difficult for the conventional retrieval systems because their feature vectors are not rotation-invariant. In this paper, to obtain rotation invariant feature vectors, in addition to invariant moments, the edge-direction histogram for a principal symmetry axis is introduced and is used to solve the bin shift problem of the histogram resulted from the rotated trademark. Performance evaluation has been carried out for a database of 300 kinds of trademarks including 20 kinds of typical trademarks which are reported to be difficult to retrieve when rotated, and the proposed scheme is proved to retrieve trademarks more efficiently, especially for the rotated trademarks, than the conventional methods.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.46 no.2
• /
• pp.165-172
• /
• 2010

This study was performed to estimate the swimming velocity of Pacific saury (Cololabis saira) migrated offshore Funka Bay of Hokkaido using an acoustic Doppler current profiler (OceanSurveyor, RDI, 153.6kHz) established in T/S Ushio-maru of Hokkaido University, in September 27, 2003. The ADCP's doppler shift revealed as the raw data that the maximum swimming velocity was measured 163.0cm/s, and its horizontal swimming speed and direction were $72.4{\pm}24.1\;cm/s$, $160.1^{\circ}{\pm}22.3^{\circ}$ while the surrounding current speed and direction were $19.6{\pm}8.4\;cm/s$, $328.1^{\circ}{\pm}45.3^{\circ}$. To calculate the actual swimming speed of Pacific saury in each bins, comparisons for each stratified bins must be made between the mean surrounding current velocity vectors, measured for each stratified bin, and its mean swimming velocity vectors, assumed by reference (threshold > -70dB) and 5dB margin among four beams of ADCP. As a result, the actual averaged swimming velocity was 88.6cm/s and the averaged 3-D swimming velocity was 91.3cm/s using the 3-D velocity vector, respectively.