• Journal of Korea Multimedia Society
• /
• v.15 no.6
• /
• pp.770-783
• /
• 2012

In this paper, we propose a novel structure recovery algorithm in the projective space using image feature points. We use normalized image feature coordinates for the numerical stability. To acquire an initial value of the structure and motion, we decompose the scaled measurement matrix using the singular value decomposition. When recovering structure and motion in projective space, we introduce matrix decomposition constraints. In the reconstruction procedure, a nonlinear iterative optimization technique is used. Experimental results showed that the proposed method provides proper accuracy and the error deviation is small.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.405-407
• /
• 1998

이 논문에서는 회전과 줌을 하는 카메라의 내부변수를 3차원 패턴 없이 주어진 영상만을 이용하여 구하는 방법을 제안한다. 먼저, 카메라의 skew를 0으로 가정하면 카메라의 내부변수가 매 영상취득 시점마다 바뀌어도 그 값들을 구할 수 있다는 것을 이론적으로 보인다. 이때 구해지는 회전 행렬을 기준 좌표계를 설정하는데 따라 달라질 뿐이다. 카메라 보정은 획득되어진 영상 사이의 투영변환행렬을 분석하여 얻어지며, 이는 회전하는 카메라에서 얻어진 두 영상 사이에는 투영변환행렬이 존재한다는 것을 이용한 것이다. 가장 일반적인 경우, 즉 카메라의 skew를 0으로 가정하면, 카메라 내부변수를 계산하기 위하여 모두 네 개의 투영변환행렬이 필요하며, 가장 단순한 카메라 모델의 경우 principal point와 aspect ratio 가 변화하지 않으면 단지 한 개의 투영거리변환이 필요하다. 합성 데이터와 실제 영상 데이터를 이용하여 제안하는 알고리듬을 시험하였다.

• Journal of Digital Contents Society
• /
• v.16 no.4
• /
• pp.605-613
• /
• 2015

This paper addresses the problem to make sparse the projection matrix in pattern recognition method. Recently, the size of computer program is often restricted in embedded systems. It is very often that developed programs include some constant data. For example, many pattern recognition programs use the projection matrix for dimension reduction. To improve the recognition performance, very high dimensional feature vectors are often extracted. In this case, the projection matrix can be very big. Recently, RSR(roated sparse regression) method[1] was proposed. This method has been proved one of the best algorithm that obtains the sparse matrix. We propose three methods to improve the RSR; outlier removal, sampling and elastic net RSR(E-RSR) in which the penalty term in RSR optimization function is replaced by that of the elastic net regression. The experimental results show that the proposed methods are very effective and improve the sparsity rate dramatically without sacrificing the recognition rate compared to the original RSR method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.5
• /
• pp.648-654
• /
• 2009

In this paper, a method called bilateral diagonal 2DLDA is proposed for face recognition. Two methods called Dia2DPCA and Dia2DLDA were suggested to reserve the correlations between the variations in the rows and columns of diagonal images. However, these methods work in the row direction of these images. A row-directional projection matrix can be obtained by calculating the between-class and within-class covariance matrices making an allowance for the column variation of alternative diagonal face images. In addition, column-directional projection matrix can be obtained by calculating the between-class and within-class covariance matrices making an allowance for the row variation in diagonal images. A bilateral projection scheme was applied using left and right multiplying projection matrices. As a result, the dimension of the feature matrix and computation time can be reduced. Experiments carried out on an ORL face database show that the proposed method with three different distance measures, namely, Frobenius, Yang and AMD, is more accurate than some methods, such as 2DPCA, B2DPCA, 2DLDA, etc.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05a
• /
• pp.571-574
• /
• 2003

본 논문은 수평 이동 정보와 회전 정보가 있는 비디오 영상을 Mellin Transform 을 이용하여 이미지를 모자이킹하는 방법을 제안한다. Mellin Transform 후 나타나는 이미지의 이동 정보와 회전정보를 이용하여 각 이미지들을 접합하기 위한 투영 행렬을 계산한다. 그리고 모자이크 생성시 나타날 수 있는 각 이미지간의 오차가 누적되는 현상을 줄이기 위한 전체적인 접합으로 투영 행렬을 추출하여 적용함으로써 누적되는 오차를 줄여 정확한 모자이크를 얻을 수 있다. 지금까지 제안된 모자이킹 기법들은 중복성 계산에 시간이 많이 소모되고 수평이동 시켜 얻어진 영상만을 다루어 이미지가 회전되었을 정확한 모자이크 결과를 얻을 수 없었다. 따라서 본 논문에서는 Mellin Transform에 기반한 투영 행렬을 이용하여 이미지가 이동하거나 회전하였을 경우에도 빠른 시간에 이미지의 중복 정보를 찾아 정확한 모자이크를 생성할 수 있는 방법을 제안한다.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.42 no.8 s.338
• /
• pp.25-32
• /
• 2005

Frequency hopping system has been adopted to many communication systems in order to overcome the inferior situation such as jamming environment. But typically its processing gain being limited, data interfered by jamming signal could not be fully recovered. This can be enhanced by combing FH system with spatial interference canceller which is a kind of active beamformer In this Paper, we proposed the compensation method of weight vector discrepancy according to the hopped frequencies and the PMBF method which is able to eliminate the inference effectively with less computational complexity. That is, the steering vector of wanted signals can be calculated from the frame without jamming signals using eigen analysis. New projection matrix extracted by the steering vector of wanted signal eliminates the interferences from the covariance matrix of received signal including wanted signal and jamming signals. This PMBF has similar performance of SINR beamformer with less computational complexity.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10a
• /
• pp.773-776
• /
• 2001

본 논문에서는 캘리브레이션 패턴을 카메라와 평행한 평면에 위치시켜 카메라의 회전 요소를 없게 하여 회전 행렬이 단위 행렬이 되게 한 상태에서 카메라를 캘리브레이션 하였다. 카메라의 투영중심에 대한 투영면과 프레임 버퍼에서의 영상 중심 위치($C_x,\;C_y$), 촛점거리(f) 등의 내부 파라메타와 세계 좌표의 원점에 대한 카메라의 상대 위치($T_x,\;T_y,\;T_z$)를 서로 다른 거리에서 획득한 두 장의 패턴 영상을 이용하여 효과적으로 계산하였다.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.22 no.1
• /
• pp.170-176
• /
• 1998

In this paper, the column space and null space of the Jacobian matrix were obtained by using the pseudo-inverse method and projection matrix. The equations of motion of the system were replaced by independent acceleration components using the null space matrix. The proposed method has the following advantages. (1) It is simple to derive the null space. (2) The efficiency is improved by getting rid of constrained force terms. (3) Neither null space updating nor coordinate partitioning method is required. The suggested algorithm is applied to a three-dimensional vehicle model to show the efficiency.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.40 no.7
• /
• pp.1257-1265
• /
• 2015

In this paper, we describe a greedy user selection scheme for multiuser multiple-input multiple-output (MIMO) systems. We propose a new metric which has significantly improved performance compared to the Frobenius norm metric. The approximation of projection matrix is applied to increase the accuracy of Frobenius norm of effective channel matrix. We analyze the computational complexity of two metrics by using flop counts, and also verify the achievable sum rate through numerical simulation. Our simulation result shows that the proposed metric can achieve the improved sum rate as the number of user antenna increases.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.8
• /
• pp.1785-1791
• /
• 2003

This paper presents the mosaic method that the video sequence with shift and rotation information after Mellin Transform. The results are used to compute the projection matrix for each image registration. So before registration, we process camera calibration in order to reduce the image warp by camera and then compute the global projection matrix for image registration for reducing errors from rut image to last image. This paper describes the mosaic method that compute duplication and movement information quickly and robust noise using projection matrix on Mellin Transform.