• ETRI Journal
• /
• v.34 no.4
• /
• pp.572-582
• /
• 2012

The discrete Gaussian-Hermite moment (DGHM) is a global feature representation method that can be applied to square images. We propose a modified DGHM (MDGHM) method and an MDGHM-based scale-invariant feature transform (MDGHM-SIFT) descriptor. In the MDGHM, we devise a movable mask to represent the local features of a non-square image. The complete set of non-square image features are then represented by the summation of all MDGHMs. We also propose to apply an accumulated MDGHM using multi-order derivatives to obtain distinguishable feature information in the third stage of the SIFT. Finally, we calculate an MDGHM-based magnitude and an MDGHM-based orientation using the accumulated MDGHM. We carry out experiments using the proposed method with six kinds of deformations. The results show that the proposed method can be applied to non-square images without any image truncation and that it significantly outperforms the matching accuracy of other SIFT algorithms.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.5
• /
• pp.87-97
• /
• 2004

In this paper, we propose an algorithm for interlaced-to-progressive conversion by the weighted summation of the information collected from spatial do-interlacing method, in which the weighted edge based line average is applied, and the temporal method in which the motion compensation is employed by using hi-directional BMA (block matching algorithm). We employed time-recursive and motion adaptive processing as motion detection is involved. Also, a median filter is used to deal with limitation of the linear summation in which only an intermediate of values being involved is determined. The main goal of the approach is to overcome the shortcomings of each of the do-interlacing techniques without significant increment of the computational complexity, and the proposed method is apt to implement in hardware for real-time processing.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.5
• /
• pp.83-91
• /
• 2001

This paper proposes the Focused beamforming to estimate the location of target residing near to the observation platform in the underwater environment. The Focused beamforming technique provides the location of target by the coherent summation of a series of incident spherical waveforms considering distinct propagation delay times at the sensor array. But due to the movement of the observation platform and the variation of the underwater environment, the shape of the sensor array is no longer to be linear but it becomes distorted as the platform moves. Thus the Focused beamforming should be peformed regarding to the geometric shape variation at each time. To estimate the target location, the artificial image plane comprised of cells is constructed, and the delays are calculated from each cell where the target could be proximity to sensors for the coherent summation. After the coherent combining, the beam pattern can be obtained through the Focused beamforming on the image plane. Futhermore to compensate the variation of the shape of the sensor array, the paper utilizes the Nth-order polynomial approximation to estimate the shape of the sensor array obeying the water pulley modeling. Simulation results show the performance of the Focused beamforming for different frequency bands of the radiated signal.

• Geophysics and Geophysical Exploration
• /
• v.21 no.2
• /
• pp.112-124
• /
• 2018

Ocean-bottom seismic survey is a seismic acquisition technique which measures data by installing 4-component receiver on the sea floor. It can produce more improved data in quality than any other acquisition techniques. In the ocean-bottom seismic survey, however, the number of receivers is limited due to high cost. Since only a small number of receivers are used for acquisition, ocean-bottom seismic data may suffer from discontinuities of events over traces, which can result in spatial aliasing. In this paper, we implemented Kirchhoff migration using mirror-imaging algorithm to improve the quality of ocean-bottom seismic image. In order to implement the mirror imaging algorithm, the seismograms should be separated into up-going and down-going wavefields and the down-going wavefield should be used for migration. In this paper, we use the P-Z summation method to separate the wavefield. Numerical examples show that the migration results using mirror imaging algorithm have wider illumination than the conventional migration, especially in the shallow layers.

• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.6
• /
• pp.494-500
• /
• 2004

This paper presents a machine vision system, which inspects the measurement of the ALC block on a real-time basis in the production process. The automatic measurement system was established with a CCD camera, an image grabber, and a personal computer without using assembled measurement equipment. Images obtained by this system was processed by an algorithm, specially designed for an enhanced measurement accuracy. For the realization of the proposed algorithm, a preprocessing method that can be applied to overcome uneven lighting environment, boundary decision method, unit length decision method in uneven condition with rocking objects, and a projection of region using pixel summation are developed. From our experimental results, we could find that the required measurement accuracy specification is sufficiently satisfied by using the proposed method.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1999.06a
• /
• pp.171-176
• /
• 1999

Template matching is one of the most often used techniques for machine vision applications to find a template of size M$\times$M or subimage in a scene image of size N$\times$N. Most template matching methods, however, require pixel operations between the template and the image under analysis resulting in high computational cost of O(M2N2). So in this thesis, we present a two stage template matching method. In the first stage, we use a novel low cost feature whose complexity is approaching O(N2) to select matching candidates. In the second stage, we use conventional template matching method to find out the exact matching point. We compare the result with other methods in terms of complexity, efficiency and performance. Proposed method was proved to have constant time complexity and to be quite invariant to noise.

• Journal of Digital Contents Society
• /
• v.8 no.3
• /
• pp.329-339
• /
• 2007

In this paper we present a facial region detection algorithm for real-time image with complex background and various illumination using spatial and temporal methods. For Detecting Human region It used summation of Edge-Difference Image between continuous image sequences. Then, Detected facial candidate region is vertically divided two objected. Non facial region is reduced using Analysis of Major Color Component. Non facial region has not available Major Color Component. And then, Background is reduced using boundary information. Finally, The Facial region is detected through horizontal, vertical projection of Images. The experiments show that the proposed algorithm can detect robustly facial region with complex background various illumination images.

• Journal of information and communication convergence engineering
• /
• v.9 no.3
• /
• pp.319-324
• /
• 2011

In computing the optical flow. Horn and Schunck's method which is a representative algorithm is based on differentiation. But it is difficult to estimate the velocity for a large displacement by this algorithm. To cope with this problem multigrid method has been proposed. In this paper, we have proposed a scaled multigrid algorithm which the initial flow for a level is calculated by the summation of the optimally scaled flow and error flow. The optimally scaled flow is the scaled expanded flow of the previous level, which can generate an estimated second image having the least RMS error with respect to the original second image, and the error flow is the flow between the estimated second image (generated by the optimally scaled flow) and the original second image. The flow for this level is then estimated using the original first and second images and the initial flow for that level. From among the various coarsest starting levels of the multigrid algorithm, we select the one that finally gives the best estimated flow. Better results were achieved using our proposed method compared with Horn and Schunck's method and a conventional multigrid algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.10
• /
• pp.4976-4994
• /
• 2018

The diffusion-based salient object detection methods have shown excellent detection results and more efficient computation in recent years. However, the current diffusion-based salient object detection methods still have disadvantage of detecting the object appearing at the image boundaries and different scales. To address the above mentioned issues, this paper proposes a multi-scale diffusion-based salient object detection algorithm with background and objectness seeds. In specific, the image is firstly over-segmented at several scales. Secondly, the background and objectness saliency of each superpixel is then calculated and fused in each scale. Thirdly, manifold ranking method is chosen to propagate the Bayessian fusion of background and objectness saliency to the whole image. Finally, the pixel-level saliency map is constructed by weighted summation of saliency values under different scales. We evaluate our salient object detection algorithm with other 24 state-of-the-art methods on four public benchmark datasets, i.e., ASD, SED1, SED2 and SOD. The results show that the proposed method performs favorably against 24 state-of-the-art salient object detection approaches in term of popular measures of PR curve and F-measure. And the visual comparison results also show that our method highlights the salient objects more effectively.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.8 no.2
• /
• pp.31-40
• /
• 2012

Face recognition refers to the process of identifying individuals based on their facial features. It has recently become one of the most popular research areas in the fields of computer vision, machine learning, and pattern recognition because it spans numerous consumer applications, such as access control, surveillance, security, credit-card verification, and criminal identification. However, illumination variation on face generally cause performance degradation of face recognition systems under practical environments. Thus, this paper proposes an novel face recognition system using a fusion approach based on local binary pattern and two-dimensional principal component analysis. To minimize illumination effects, the face image undergoes the local binary pattern operation, and the resultant image are divided into two sub-images. Then, two-dimensional principal component analysis algorithm is separately applied to each sub-images. The individual scores obtained from two sub-images are integrated using a weighted-summation rule, and the fused-score is utilized to classify the unknown user. The performance evaluation of the proposed system was performed using the Yale B database and CMU-PIE database, and the proposed method shows the better recognition results in comparison with existing face recognition techniques.