• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2014.10a
• /
• pp.572-574
• /
• 2014

Researches on three-dimensional multimedia has been performed actively in recent years. Virtual sound technology corresponding to virtual image should be provided to implement 3D multimedia with high quality. Head-related transfer function (HRTF) plays a key role in this research area. HRTFs measured in changing azimuth, elevation, and distance for each and every subject is necessary for ideal solution. However, it is practically impossible to measure all subjects' HRTFs, so various HRTF databases have been built by many researchers. Because HRTF displays quite different aspects from subject to subject, HRTF of dummy head has been used for generic usage. However, mannequin's HRTF showed much worse performance comparing with individual case so this solution should be improved. From previous work, standardization of HRTF based on tensor-singular value decomposition method has been proposed. For effective extraction of standard HRTF, three different decomposition methods are compared in this paper.

• Journal of the Optical Society of Korea
• /
• v.8 no.3
• /
• pp.108-114
• /
• 2004

To allow multiple viewers to see the correct perspective and to provide a singh1 viewer with motion parallax cues during head movement, more than two views arc needed. Since it is prohibitive to acquire, process, and transmit a continuum of views, it. would preferable to acquire only a minimal set of views and to synthesize intermediate images. This paper presents how to synthesize the intermediate images using irregular quadtree decomposition and compares the proposed methods with full-search block matching. The image at the middle viewpoint between both viewpoints is synthesized and yields a 32.8 ㏈ peak-signal-to-noise ratio (PSNR) which is 2.8 ㏈ high and has a running time 30% of that for conventional full-search block matching.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.12
• /
• pp.81-90
• /
• 1994

This paper proposes a motphological fast reconstructive filter (FRF) using up/down sampling techniques for reconstructive opening and closing, and a parallel structure for fast multiresolution decomposition. Compuer simulation shows that, compared with the conventional RF, the proposed FRF can reduce the processing time up to 8 times while it maintains a similar performance in reconstructed shapes. Further reduction in the decomposition time achieved by the paralellized algorithm combined with the FRF, which can be applied in areas such as defect detection, image segmentation, pattern recognition, etc.

• Proceedings of the Korean Society Of Semiconductor Equipment Technology
• /
• 2004.05a
• /
• pp.217-222
• /
• 2004

This paper focuses on lossy medical image compression methods for medical images that operate on two-dimensional(2D) integer wavelet transform. We offer an application of the Set Partitioning in Hierarchical Trees(SPIHT) algorithm to medical images, using a 2D wavelet decomposition and a 2D spatial dependence tree. The wavelet decomposition is accomplished with integer wavelet filters implemented with the lifting method, where careful scaling and truncations keep the integer precision small and the transform unitary. We have tested our encoder on medical images using different integer filters. Results show that our algorithm with certain filters performs as well and is sometimes better lossy coding using 2D integer wavelet transforms on medical images.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.11
• /
• pp.72-82
• /
• 1996

In this paper, newly developed algorithm for enhancing images corrupted by white gaussian noise is proposed. In the method proposed here, image is subdivided into a number of subblocks, and each block is separated into cimponents corresponding to signal and noise subspaces, respectively through the signal subspace method. A clean signal is then estimated form the signal subspace by the adaptive wiener filtering. The decomposition of noisy signal into noise and signal subspaces in is implemented by eigendecomposition of covariance matrix for noisy image, and by performing blockwise KLT (karhunen loeve transformation) using eigenvector. To reduce the perceptual noise level and distortion, wiener filtering is implementd by adaptively adjusting noise level according to activity characteristics of given block. Simulation results show the effectiveness of proposed method. In particular, edge bluring effects are reduced compared to the previous methods.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.395-398
• /
• 2003

Two component analysis methods based on the fuzzy relational calculus are proposed in the setting of the ordered structure. First component analysis is based on a decomposition of fuzzy relation into fuzzy bases, using gradient method. Second one is a component analysis based on the eigen fuzzy sets of fuzzy relation. Through experiments using the test images extracted from SIDBA and View Sphere Database, the effectiveness of the proposed component analysis methods is confirmed. Furthermore, improvements of the image compression/reconstruction and image retrieval based on ordered structure are also indicated.

• Journal of Electrical Engineering and Technology
• /
• v.4 no.4
• /
• pp.552-558
• /
• 2009

We proposed a semi-blind digital image watermarking technique for copyright protection. The proposed algorithm embedded a binary sequence watermark into significant wavelet coefficients by using a quantization method. The main idea of the quantization method was to quantize a middle coefficient of the triplet of a significant wavelet coefficient according to the watermark's value. Unlike an existing algorithm, which used a random location table to find a coefficient in which the watermark bit will be embedded: the proposed algorithm used quad-tree decomposition to find a significant wavelet coefficient for embedding. For watermark detection, an original host image was not required. Thanks to the usage of significant wavelet coefficients, the proposed algorithm improved the correlation value, up to 0.43, in comparison with the existing algorithm.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.11
• /
• pp.2750-2767
• /
• 1996

In this papr, a new lossless compression method is presented based on the Binary Adaptive Arithmetic Coder(BAAC). A simple unbalanced binary tree is created by recursively dividing the BAAC unit interval into two probability sub-inervals. On the tree the More Probable Predicted Value(MPPV) and Less Probable Predicated Value(LPPV) estimated by local statistics of the image pixels are arranged in decreasing order. The BAAC or Huffman coder is thus applied to the branches of the tree. The proposed method allows the coder be directly applied to the full bit-plane medical image without a decomposition of the full bit-planes into a series of binary bit-planes. The use of the full bit model template improves the compresion ratio. In addition, a fast computation for adjusting the interval is possible since a simple arithmetic operation based on probability interval estimation state machine is used for interval sub-division within the BAAC unit interval.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.5
• /
• pp.1141-1154
• /
• 1996

The para-unitary (PU) M band filter bank, which can beused for M band decomposition, has many useful properties. In this ppaer, attempts have been made to design and and analyze the linear phase para-unitary (LPPU) M band filter bank, which is appropriate to the image coding application. First, we derive a unified coding gain in terms of the correlation in the band, as well as the energy compaction. And M band filter bank has been designed, maximizing the new coding gain. Then, we analyze the image coding performance of the LPPU M band filter bank, such as the energy compaction, the correlation in the band and the entropy. From the analysis, it is shown that the coding gain for LPPU M band filter bank improves, and the coding gain for the LPPU 4 band filter bank approaches very closely to that for LPPU 8 band filter bank, as the length of the filter increases. This fact is also verified by the coding results on the real images.

• Proceedings of the KIEE Conference
• /
• 1989.11a
• /
• pp.415-420
• /
• 1989

This thesis deals with range finding using one camera and laser pointer. Range finding will be used further recognition of the image, that is, range image which allows further segmentation of the scene. In the first step, camera modeling is performed by camera calibration which executes least square fit. Least square fit uses the method of sigular value decomposition. And perspective transform of camera is obtained. Lastly range finding is performed by triangulation principle. The result of this algorithm are displayed.