运动估计算法分析

对常用的运动补偿图像算法（如全搜索法、三步法、共轭方向搜索法、两维对数下降法等）从各个方面进行了比较,最终选出最快、最准确的算法-两步搜索法。     

Three-dimensional motion estimation of objects for video coding

Three-dimensional (3-D) motion estimation is applied to the problem of motion compensation for video coding. We suppose that the video sequence consists of the perspective projections of a collection of rigid bodies which undergo a rototranslational motion. Motion compensation can be performed on the sequence once the shape of the objects and the motion parameters are determined. We show that the motion equations of a rigid body can be formulated as a nonlinear dynamic system whose state is represented by the motion parameters and by the scaled depths of the object feature points. An extended Kalman filter is used to estimate both the motion and the object shape parameters simultaneously. The inclusion of the shape parameters in the estimation procedure adds a set of constraints to the filter equations that appear to be essential for reliable motion estimation. Our experiments show that the proposed approach gives two advantages. First, the filter can give more reliable estimates in the presence of measurement noise in comparison with other motion estimators that separately compute motion and structure. Second, the filter can efficiently track abrupt motion changes. Moreover, the structure imposed by the model implies that the reconstructed motion is very natural as opposed to more common block-based schemes. Also, the parameterization of the model allows for a very efficient coding of the motion information     

Level-set-based motion estimation algorithm for multiple reference frame motion estimation

Motion estimation (ME) has a variety of applications in image processing, pattern recognition, target tracking, and video compression. In modern video compression standards such as H.264/AVC and HEVC, multiple reference frame ME (MRFME) is adopted to reduce the temporal redundancy between successive frames in a video sequence. In MRFME, the motion search process is conducted using additional reference frames, thereby obtaining better prediction signal as compared to single reference frame ME (SRFME). However, its high computational complexity makes it difficult to be utilized in real-world applications. In order to reduce the computational complexity of MRFME, this paper proposes a level-set-based ME algorithm (LSME) without any penalty in the rate-distortion (RD) performance. First, the proposed algorithm partitions the motion search space into multiple level sets based on a rate constraint. The proposed algorithm then controls the ME process on the basis of the predetermined level sets. Experimental results show that the proposed algorithm reduces the ME time by up to 83.46% as compared to the conventional full search (FS) algorithm.     

Spatiotemporal motion boundary detection and motion boundary velocity estimation for tracking moving objects with a moving camera: a level sets PDEs approach with concurrent camera motion compensation

The purpose of this study is to investigate a method of tracking moving objects with a moving camera. This method estimates simultaneously the motion induced by camera movement. The problem is formulated as a Bayesian motion-based partitioning problem in the spatiotemporal domain of the image quence. An energy functional is derived from the Bayesian formulation. The Euler-Lagrange descent equations determine imultaneously an estimate of the image motion field induced by camera motion and an estimate of the spatiotemporal motion undary surface. The Euler-Lagrange equation corresponding to the surface is expressed as a level-set partial differential equation for topology independence and numerically stable implementation. The method can be initialized simply and can track multiple objects with nonsimultaneous motions. Velocities on motion boundaries can be estimated from geometrical properties of the motion boundary. Several examples of experimental verification are given using synthetic and real-image sequences.     

Fast algorithms for the estimation of motion vectors

In this correspondence, a fast motion estimation algorithm, based on the successive elimination algorithm (SEA) of Li and Salari (1995), is studied. This fast motion estimation algorithm finds the same displacement vectors as the exhaustive search algorithm with a reduced computational load. A modified fast motion estimation algorithm introducing negligible distortion into a transform coder, but providing for a further computational load reduction, is developed. Implementation issues are also discussed and compared. Results show that the number of searching operations can be reduced dramatically with the help of fast motion estimation algorithms.     

A new technique for velocity estimation of large moving objects

A technique for motion estimation of large moving objects is presented. This technique is based on analyzing the Hartley transform spectrum of the image sequence directly, instead of using it to compute other transforms. In this technique, the fast Hartley transform (FHT) is applied to the image sequence and followed by a peak detection procedure. The location of the peak is related to the velocity of the moving object. Dividing the temporal frequency (f_p) corresponding to the detected peak by the corresponding spatial frequency (k_p) gives the velocity of the moving object. The Fourier spectrum for a spatial frequency of k_p is then computed. This is followed by a peak detection of the Fourier spectrum to validate the previous results and find the direction of the velocity of the moving object. This method is faster than other techniques based on the Fourier transform     

Kernel estimation from salient structure for robust motion deblurring

Blind image deblurring algorithms have been improving steadily in the past years. Most state-of-the-art algorithms, however, still cannot perform perfectly in challenging cases, especially in large blur setting. In this paper, we focus on how to estimate a good blur kernel from a single blurred image based on the image structure. We found that image details caused by blur could adversely affect the kernel estimation, especially when the blur kernel is large. One effective way to remove these details is to apply image denoising model based on the total variation (TV). First, we developed a novel method for computing image structures based on the TV model, such that the structures undermining the kernel estimation will be removed. Second, we applied a gradient selection method to mitigate the possible adverse effect of salient edges and improve the robustness of kernel estimation. Third, we proposed a novel kernel estimation method, which is capable of removing noise and preserving the continuity in the kernel. Finally, we developed an adaptive weighted spatial prior to preserve sharp edges in latent image restoration. Extensive experiments testify to the effectiveness of our method on various kinds of challenging examples.     

Efficient multilevel successive elimination algorithms for block matching motion estimation

The authors present fast algorithms to reduce the computations of block matching algorithms, for motion estimation in video coding. Efficient multilevel successive elimination algorithms are based on the multilevel successive elimination. Efficient multilevel successive elimination algorithms consist of four algorithms. The first algorithm is given by the sum of absolute difference between the sum norms of sub-blocks in a multilevel successive elimination algorithm (MSEA) using the partial distortion elimination technique. By using the first algorithm, computations of MSEA can be reduced further. In the second algorithm, the sum of absolute difference (SAD) is calculated adaptively from large value to small value according to the absolute difference values between pixels of blocks. By using the second algorithm, the partial distortion elimination in the SAD calculation can occur early, therefore the computations of MSEA can be reduced. The second algorithm is useful not only with MSEA, but also with all kinds of block matching algorithms. In the third algorithm, the elimination level of the MSEA can be estimated. Accordingly, the computations of the MSEA related to the level lower than the estimated level can be reduced. The fourth algorithm is, first of all, to search the motion vector over the half sampled search points. At the second search, the authors search the unsampled search points around the tested search points where the motion vector may exist from the first search results. The motion estimation accuracy of the fourth algorithm is nearly 100% and the computations can be greatly reduced.     

Fast multiresolution motion estimation algorithms for wavelet-based scalable video coding

Motion estimation and compensation in wavelet domain have received much attention recently. To overcome the inefficiency of motion estimation in critically sampled wavelet domain, the low-band-shift (LBS) method and the complete-to-overcomplete discrete wavelet transform (CODWT) method are proposed for motion estimation in shift-invariant wavelet domain. However, a major disadvantage of these methods is the computational complexity. Although the CODWT method has reduced the computational complexity by skipping the inverse wavelet transform and making the direct link between the critically sampled subbands and the shift-invariant subbands, the full search algorithm (FSA) increases it. In this paper, we proposed two fast multiresolution motion estimation algorithms in shift-invariant wavelet domain: one is the wavelet matching error characteristic based partial distortion search (WMEC-PDS) algorithm, which improves computational efficiency of conventional partial distortion search algorithms while keeping the same estimate accuracy as the FSA; another is the anisotropic double cross search (ADCS) algorithm using multiresolution-spatio-temporal context, which provides a significantly computational load reduction while only introducing negligible distortion compared with the FSA. Due to the multiresolution nature, both the proposed approaches can be applied to wavelet-based scalable video coding. Experimental results show the superiority of the proposed fast motion estimation algorithms against other fast algorithms in terms of speed-up and quality.     

Bayesian motion estimation for temporally recursive noise reduction in X-ray fluoroscopy

This paper develops a Bayesian motion estimation algorithm for motion-compensated temporally recursive filtering of moving low-dose X-ray images (X-ray fluoroscopy). These images often exhibit a very low signal-to-noise ratio. The described motion estimation algorithm is made robust against noise by spatial and temporal regularization. A priori expectations about the spatial and temporal smoothness of the motion vector field are expressed by a generalized Gauss-Markov random field. The advantage of using a generalized Gauss-Markov random field is that, apart from smoothness, it also captures motion edges without requiring an edge detection threshold. The costs of edges are controlled by a single parameter, by means of which the influence of the regularization can be tuned from a median-filter-like behaviour to a linear-filter-like one.     

两种实时、低码率视频编码的运动估计搜索算法

先提出实时、低码率视频编码对运动估计搜索算法在运算时间和准确度上的要求，总结归纳了钻石搜索法和两步法的具体算法，通过与几种经典算法作比较，说明这两种算法更优，更适合用于实时，低码率视频编码，最后，以H．263的实现为例，说明了钻石搜索算法的具体实现。     

Data reuse algorithm for multiple reference frame motion estimation

A data reuse algorithm for multiple reference frame motion estimation is described. The proposed algorithm reduces memory access by modifying the reference frame search order and search centre such that the likelihood of data reuse is increased. Experimental results show that the algorithm reduces memory access by 15-30% compared to the conventional fast reference frame selection algorithm, while maintaining similar bit rate and PSNR     

一种新的H.264多参考帧快速搜索算法

H.264/AVC是目前最新的视频编码国际标准.在H.264/AVC中,由于采用了多参考帧,运动估计部分的复杂度大大增加了.虽然JVT采用的快速运动估计(FME)提案能极大地提高搜索速度,但只是加快了一帧的搜索速度.本文在FME基础上提出了一种新的多参考帧快速搜索算法(FMRSA),它利用当前块周围的块信息来预测将要搜索的参考帧的范围,并在具体搜索过程中运用提前中断的方法加快整个搜索过程.实验证明本算法比标准快速搜索法搜索5个参考帧能在PSNR降低不超过0.05dB、码率增加不超过2.32%的情况下节省至少56.5%的时间.     

Preliminary study on helical CT algorithms for patient motion estimation and compensation

Helical computed tomography (helical/spiral CT) has replaced conventional CT in many clinical applications. In current helical CT, a patient is assumed to be rigid and motionless during scanning and planar projection sets are produced from raw data via longitudinal interpolation. However, rigid patient: motion is a problem in some cases (such as in the skull base and temporal bone imaging). Motion artifacts thus generated in reconstructed images can prevent accurate diagnosis. Modeling a uniform translational movement, the authors address how patient motion is ascertained and how it may be compensated. First, mismatch between adjacent fan-beam projections of the same orientation is determined via classical correlation, which is approximately proportional to the patient displacement projected onto an axis orthogonal to the central ray of the involved fan-beam. Then, the patient motion vector (the patient displacement per gantry rotation) is estimated from its projections using a least-square-root method. To suppress motion artifacts, adaptive interpolation algorithms are developed that synthesize full-scan and half-scan planar projection data sets, respectively. In the adaptive scheme, the interpolation is performed along inclined paths dependent upon the patient motion vector. The simulation results show that the patient motion vector can be accurately and reliably estimated using the authors' correlation and least-square-root algorithm, patient motion artifacts can be effectively suppressed via adaptive interpolation, and adaptive half-scan interpolation is advantageous compared with its full-scan counterpart in terms of high contrast image resolution.     

TNRAC: a system for tracking multiple moving non-rigid objects using an active camera

This paper presents an enhanced method for detecting and tracking moving objects from an active camera image sequence using active contour models. Such enhancements are: (1) Using best block matching instead of the conventional block matching technique to locate the promising feature points in the current frame. (2) Developing a new algorithm for rejecting bad feature points that affect negatively the background compensation process. (3) Applying a modified region growing and labeling algorithm to extract moving objects then applying active contour models around these moving objects in order to take into account the object inside this contour. Various factors are studied to measure the system performance and limitations. A comparison of TNRAC and the conventional approach is conducted using both rigid and non-rigid objects in both indoor and outdoor scenes.     

A multiple antenna wireless testbed for the validation of DoA estimation algorithms

In this paper, a wireless testbed with a multiple antenna receiver is described. It comprises a rotating four-element antenna array connected to a quad radio frequency (RF) front-end, and a channel acquisition board equipped with a field programmable gate array (FPGA). The algorithms can be implemented directly on the FPGA, but they can also be tested via simulation, e.g., with Matlab, since the acquired data can be transferred from the board memory to a personal computer (PC). Moreover, the algorithm implementation on the FPGA can be done by exploiting the System Generator for DSP Xilinx tool that allows the algorithm synthesis from a Simulink block diagram. These features make the testbed useful for rapid prototyping. In particular, the presence of the rotating antenna array enables the analysis of direction of arrival (DoA) estimation techniques. We report the main results from the experimental measures conducted to characterize the hardware non idealities. We then describe a DoA estimation algorithm that has been used to compare real and simulated results. It is shown that the results are in good agreement with the simulations, also when the effect of non isotropic antenna gains and/or phase noise originated from non co-phased RF front-ends becomes considerable.     

Fast motion estimation for H.264

Several specific features have been incorporated into Motion estimation (ME) in H.264 coding standard to improve its coding efficiency. However, they result in very high computational load. In this paper, a fast ME algorithm is proposed to reduce the computational complexity. First, a mode discriminant method is used to free the encoder from checking the small block size modes in homogeneous regions. Second, a condensed hierarchical block matching method and a spatial neighbor searching scheme are employed to find the best full-pixel motion vector. Finally, direction-based selection rule is utilized to reduce the searching range in sub-pixel ME process. Experimental results on commonly used QCIF and CIF format test sequences have shown that the proposed algorithm achieves a reduction of 88% ME process time on average, while incurring only 0.033 dB loss in PSNR and 0.50% increment on the total bit rate compared with that of exhaustive ME process, which is a default approach adopted in the JM reference software.     

Recursive T-matrix methods for scattering from multiple dielectricand metallic objects

We present an efficient, stable, recursive T-matrix algorithm to calculate the scattered field from a heterogeneous collection of spatially separated objects. The algorithm is based on the use of higher order multipole expansions than those typically employed in recursive T-matrix techniques. The use of these expansions introduces instability in the recursions developed by Chew (1990) and by Wang and Chew (1990), specifically in the case of near-field computations. By modifying the original recursive algorithm to avoid these instabilities, we arrive at a flexible and efficient forward solver appropriate for a variety of scattering calculations. The algorithm can be applied when the objects are dielectric, metallic, or a mixture of both. We verify this method for cases where the scatterers are electrically small (fraction of a wavelength) or relatively large (12λ). While developed for near-field calculation, this approach is applicable for far-field problems as well. Finally, we demonstrate that the computational complexity of this approach compares favorably with comparable recursive algorithms     

运动物体大气扰流的可视化光学监测方法

建立了对运动物体在大气层内运动引发的大气扰流进行可视化光学监测的方法,包括大气扰流光传输方法、大气扰流光偏折监测方法,以及高精度扰动检出方法三个部分。首先,在分析大气扰流引发光线偏折机理的基础上,研究了大气扰流区光线折射率、折射率梯度和光线偏折传输的计算方法;其次,基于背景纹影成像原理的监测系统各参数对监测效能影响的分析方法;最后,利用“整像素搜索+亚像素定位”的高精度扰动检出方法,分别建立了这三种方法的数学模型并开展了仿真分析。结果表明:该方法为可视化监测航天飞行器返回大气层着陆减速、降落伞动力展开、客机超音速飞行等过程中与大气的相互作用,提供了一种重要的方法与途径,监测结果可为优化气动外形提供数据支撑。