• Journal of KIISE:Software and Applications
• /
• v.31 no.2
• /
• pp.245-254
• /
• 2004

This paper deals with the problem of estimating structure and motion from long continuous image sequences, applying the Expectation Maximization algorithm based on extended Kalman smoother to impose the time-continuity of the motion parameters. By repeatedly estimating the state transition matrix of the dynamic equation and the parameters of noise processes in the dynamic and measurement equations, this optimization gives the maximum likelihood estimates of the motion and structure parameters. Practically, this research is essential for dealing with a long video-rate image sequence with partially unknown system equation and noise. The algorithm is implemented and tested for a real image sequence.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.13 no.5
• /
• pp.430-442
• /
• 2013

This paper presents a fast multi-reference frame integer motion estimator for H.264/AVC. The proposed system uses the previously proposed fast multi-reference frame algorithm. The previously proposed algorithm executes a full search area motion estimation in reference frames 0 and 1. After that, the search areas of motion estimation in reference frames 2, 3 and 4 are minimized by a linear relationship between the motion vector and the distances from the current frame to the reference frames. For hardware implementation, the modified algorithm optimizes the search area, reduces the overlapping search area and modifies a division equation. Because the search area is reduced, the amount of computation is reduced by 58.7%. In experimental results, the modified algorithm shows an increase of bit-rate in 0.36% when compared with the five reference frame standard. The pipeline structure and the memory controller are also adopted for real-time video encoding. The proposed system is implemented using 0.13 um CMOS technology, and the gate count is 1089K with 6.50 KB of internal SRAM. It can encode a Full HD video ($1920{\times}1080P@30Hz$) in real-time at a 135 MHz clock speed with 5 reference frames.

• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.5
• /
• pp.471-478
• /
• 2012

A foot motion is estimated using an inertial sensor unit, which is installed on a shoe. The inertial sensor unit consists of 3 axis accelerometer and 3 axis gyroscopes. Attitude and position of a foot are estimated using an inertial navigation algorithm. To increase estimation performance, a smoother is used, where the smoother employs a forward and backward filter structure. An indirect Kalman filter is used as a forward filter and backward filter. A new combining algorithm for the smoother is proposed to combine a forward indirect Kalman filter and a backward indirect Kalman filter. Through experiments, the estimation performance of the proposed smoother is verified.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.887-890
• /
• 1998

We present a novel multi-resolution block matching algorithm (BMA) for fast motion estimation. At the coarsest level, a full search BMA (FSBMA) is performed for searching complex or random motion. Concurrently, spatial correlation of motion vector (MV) field is used for searching continuous motion. Here we present an efficient method for searching full resolution MVs without MV decimation even at the coarsest leve. After the coarsest level search, two or three initial MV candidates are chosen for the next level. At the further levels, the MV candidates are refined within much smaller search areas. Simulation results show that in comparison with FSBMA, the proposed BMA achieves a speed-up factor over 710 with minor PSNR degradation of 0.2dB at most, under a normal MPEG2 coding environment. Furthermore, our scheme is also suitable for hardware implementation due to regular data-flow.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.46 no.11
• /
• pp.66-76
• /
• 2009

The motion estimation in video encoding system occupies the biggest part. So, we require the motion estimator with efficient structure for real-time operation. And for motion estimator's implementation, it is desired to design hardware module of an exclusive use that perform the encoding process at high speed. This paper proposes motion estimation detection block(MED), 41 SADs(Sum of Absolute Difference) calculation block, minimum SAD calculation and motion vector generation block based on parallel processing. The parallel processing can reduce effectively the amount of the operation. The minimum SAD calculation and MED block uses the pre-computation technique for reducing switching activity of the input signal. It results in high-speed operation. The MED and 41 SADs calculation blocks are composed of adder tree which causes the problem of critical path. So, the structure of adder tree has changed the most commonly used ripple carry adder(RCA) with carry skip adder(CSA). It enables adder tree to operate at high speed. In addition, as we enabled to easily control key variables such as control signal of search range from the outside, the efficiency of hardware structure increased. Simulation and FPGA verification results show that the delay of MED block generating the critical path at the motion estimator is reduced about 19.89% than the conventional strukcture.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.33A no.3
• /
• pp.163-175
• /
• 1996

This paper invetigates estimation methods of mapping parameters in object-oriented coding. In this paper, we propose a fast parameter estimation method with its performance similar to that of the conventional methods. We employ hierarchical structure in difference images to redcue the computational complexity and also combine conventional six- and eight-mapping parameter estimation methods to compensate for the performance degradation caused by employment of hierarchical structure. Computer simulation shows that the proposed mehtod gives results similar to conventional methods with greatly reduced computational complexity.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.10a
• /
• pp.822-824
• /
• 2018

Estimation of 3-D objects from 2-D images is inherently performed by either motion or scene features methods as it has been described in different literatures. Structure from motion as a method employed in this study uses calibrated camera and reconstructed 3-D points from the structure of the scene for reliable and precise estimates. In this study we construct 3-D shapes using color pixels and reconstructed 3-D points to determine observable differences for the constructed 3-D images. The estimation using reconstructed 3-D points indicates that the sphere is recovered by the use of scale factor due to its known size while the one obtained by using color pixels has look similar to the former but different in the scales of the axes.

• Journal of the Society of Naval Architects of Korea
• /
• v.35 no.2
• /
• pp.20-28
• /
• 1998

To accurately predict the motion of a submergible, the nonlinear structure of dynamic model should be selected and corresponding parameters should be estimated using model test. Providing the model structure, only the values of parameters are unknown and the estimation can thus be formulated as a standard least square problem. Unfortunately, the nonlinear model structure of submersibles is rarely known a prior and method of model structure determination from measurement data of model test should be developed and included as a vital part of the estimation procedure. In this study, the well-known linear least square algorithm for the analysis of model tests and a way to measure the goodness are reviewed, and the identification algorithm based on an orthogonal decomposition method of Gram-Schmidt is extended to combine structure selection and maneuvering coefficients estimation in a very simple and efficient manner. Finally, the efficiency of this algorithm is verified by using simulation and applying to the analysis of model test of a submerged body. As a result, it was verified that this combined method might be very erective in selecting the structure of dynamic model estimating the maneuvering coefficients from measurement fiat of model test.

• Journal of Computational Design and Engineering
• /
• v.1 no.1
• /
• pp.1-12
• /
• 2014

In this paper, we propose a new method of reconstructing the hand models for individuals, which include the link structure models, the homologous skin surface models and the homologous tetrahedral mesh models in a reference posture. As for the link structure model, the local coordinate system related to each link consists of the joint rotation center and the axes of joint rotation, which can be estimated based on the trajectories of optimal markers on the relative skin surface region of the subject obtained from the motion capture system. The skin surface model is defined as a three-dimensional triangular mesh, obtained by deforming a template mesh so as to fit the landmark vertices to the relative marker positions obtained motion capture system. In this process, anatomical dimensions for the subject, manually measured by a caliper, are also used as the deformation constraints.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.40 no.5
• /
• pp.347-356
• /
• 2003

Camera pose and scene geometry estimation from video sequences is widely used in various areas such as image composition. Structure and motion recovery based on the auto calibration algorithm can insert synthetic 3D objects in real but un modeled scenes and create their views from the camera positions. However, most previous methods require bundle adjustment or non linear minimization process [or more precise results. This paper presents a new auto' calibration algorithm for video sequence based on two steps: the one is key frame selection, and the other removes the key frame with inaccurate camera matrix based on an absolute quadric estimation by LMedS. In the experimental results, we have demonstrated that the proposed method can achieve a precise camera pose estimation and scene geometry recovery without bundle adjustment. In addition, virtual objects have been inserted in the real images by using the camera trajectories.