一种改进遗传搜索块匹配运动估计算法

运动估计是帧间视频编码中的关键技术,但现有的快速搜索算法中大都是次优算法,且易陷于局部极小点。针对此问题,提出了一种改进型遗传算法应用于块运动估计中的遗传搜索块匹配运动估计算法(MGSAME)。该方法把块运动向量作为遗传染色体,经过杂交、变异等操作,以便得到全局意义上的最优解,并与经典的全局搜索法、三步搜索法和传统遗传算法(SGA)进行了比较。实验结果显示,该算法不仅有效地解决了局部极小问题,而且计算量也较少。     

基于时空相关性的自适应运动估计快速算法

针对视频编码中的核心技术运动估计,提出一种基于运动矢量特性的运动估计快速算法。算法分析视频序列运动矢量的特性,对静止块设定自适应阈值直接终止搜索,自适应选择搜索起点。采用2种模式进行搜索,依据当前块与邻近块的空间相关性大小,即局部运动性是高还是低,自适应地选择搜索模式和搜索终止门限。实验结果表明,该算法的搜索速度和搜索精度优于现有的快速运动估计算法,搜索精度非常接近于全搜索法。     

基于改进型C-1BT变换的快速全局运动估计

针对存在单一运动目标视频序列的全局运动估计问题,提出一种结合C-1BT变换和自适应十字搜索法(ARPS)的快速块匹配运动估计方法。采用简单的预处理,提高灰度突然变化和存在局部运动目标时的全局运动估计的鲁棒性;简化传统C-1BT变换算法中的核函数,降低计算负担;考虑到实际视频中存在大量的零运动矢量(ZMV)区域和视频的空间相关性,改进原有的ARPS搜索法。MATLAB仿真结果证明,提出的算法在保证精度的同时极大地减少了搜索点数。     

Fast global motion estimation based on improved C-1BT transformation

针对存在单一运动目标视频序列的全局运动估计问题, 提出一种结合C-1BT变换和自适应十字搜索法ARPS的快速块匹配运动估计方法。采用简单的预处理, 提高灰度突然变化和存在局部运动目标时的全局运动估计的鲁棒性; 简化传统C-1BT变换算法中的核函数, 降低计算负担; 考虑到实际视频中存在大量的零运动矢量ZMV区域和视频的空间相关性, 改进原有的ARPS搜索法。MATLAB仿真结果证明, 提出的算法在保证精度的同时极大地减少了搜索点数。     

基于免疫克隆选择的块匹配运动估计

运动估计是视频压缩编码中的关键技术.从运动矢量的特点出发,采用搜索点预测、Gray码编码以及有效的迭代终止准则等策略,提出了基于免疫克隆选择的块匹配运动估计.该方法将块匹配运动估计问题的性质与免疫克隆选择算法所具有的全局搜索特性、解的多样性和不易早熟的特点相融合,在能够获得接近全搜索方法所得到的平均峰值信噪比的前提下,使得平均搜索点数大为降低.仿真实验结果表明,在大多数序列上,该算法都比已有的快速搜索算法具有更高的性能和更少的平均搜索点数.同时,该算法适用面广,对大运动和小运动序列都能得到较好的效果.     

一种改进的快速全局运动估计算法

结合两步法与传统梯度下降算法,提出一种改进的快速全局运动估计算法。采用稀疏抽样的MSEA快速块匹配算法估计局部运动矢量,使用迭代最小二乘法粗估计全局运动参数并排除外点(前景宏块),在排除外点的采样宏块集上选取特征像素,以上述两步法的全局运动估计参数为初始值,利用LM梯度下降算法对全局运动参数进行优化。实验结果表明,改进算法的估计速度达到11.42 ms/f,比FFRGMET算法快1.3倍,具有更高的全局运动估计精度。     

一种快速分类搜索运动估计新方法

在视频编码中 ,基于块的运动估计算法被广泛应用 .在保证估计质量的前提下 ,为了降低运动估计算法的搜索次数 ,提出了一种对于不同类型的块采用不同的搜索范围和搜索步骤的分类快速搜索 (CFS)运动估计新算法 .该算法首先对块进行分类 ,然后确定其搜索范围和搜索步骤 ,在应用分类搜索法时 ,根据运动矢量的中心偏置特性 ,将第 1步和第 2步的搜索窗采用 5× 5的窗口 ,第 3步采用 3× 3的窗口 .结果表明 ,该分类快速搜索新算法在运动矢量的估计质量上 ,明显优于传统三步搜索法 ,且搜索次数与传统三步搜索法相比 ,降低了 2 3% ,与全搜索法相比 ,降低了 91% .实验结果证明 ,该算法尤其适用于快速运动、复杂运动序列的运动估计 .与传统的全搜索法和三步搜索法相比 ,其更适合于用硬件实现 .     

一种基于新型遗传算法的块运动估计算法

提出了一种基于新型遗传算法的块匹配运动估计算法。该算法把块运动向量作为遗传染色体,经过选择变异等操作,将随机搜索与特定目标搜索相结合,解决了以往快速搜索算法易陷于局部最优的问题,同时该算法中所带的模糊评价函数使得对个体的评价更合理、客观,该算法还将运动矢量空间偏置特性用于初始种群的选取,进一步提高了算法性能。实验结果表明,该算法性能上接近于FSA,速度却接近于TSS。     

基于Hilbert抽取的图像块运动估计算法

在图像块运动估计算法中,自适应抽取图像像素关键点是块匹配成功的关键因素.针对传统自适应抽取像素点算法主要针对局部图像区域的问题,通过利用Hilbert曲线来抽取全局区域的像素点,并把它与十字架搜索模板相结合,提出了一个改进的快速块匹配算法.仿真试验结果表明,该算法简单有效,可以在低比特率视频编码中得到很好的应用.     

一种新的基于H.264/AVC的快速块运动估计算法

为了解决目前快速块运动估计搜索算法容易使系统陷入局部最优陷阱这一问题,本文提出了一种新的基于H.264/AVC的快速块运动估计算法.简要介绍了运动估计算法和部分失真消除算法,给出了基于空间相关性的搜索范围修正方法,并提出了基于PDE算法的最大局部匹配误差算法.通过实验数据测试表明,该算法在缩小搜索点数,提高编码效率上具有较大的优势.     

Simple and efficient motion estimation algorithm by continuum search

Motion estimation plays a vital role in reducing temporal correlation in video codecs but it requires high computational complexity. Different algorithms have tried to reduce this complexity. However these reduced-complexity routines are not as regular as the full search algorithm (FSA). Also, regularity of an algorithm is very important in order to have a hardware implementation of that algorithm even if it leads to more complexity burden. The goal of this paper is to develop an efficient and regular algorithm which mimics FSA by searching a small area exhaustively. Our proposed algorithm is designed based on two observations. The first observation is that the motion vector of a block falls within a specific rectangular area designated by the prediction vectors. The second observation is that in most cases, this rectangular area is smaller than one fourth of the FSA’s search area. Therefore, the search area of the proposed method is adaptively found for each block of a frame. To find the search area, the temporal and spatial correlations among motion vectors of blocks are exploited. Based on these correlations, a rectangular search area is determined and the best matching block in this area is selected. The proposed algorithm is similar to FSA in terms of regularity but requires less computational complexity due to its smaller search area. Also, the suggested algorithm is as simple as FSA in terms of implementation and is comparable with many of the existing fast search algorithms. Simulation results show the claimed performance and efficiency of the algorithm.     

基于改进遗传算法的稀疏重构算法

稀疏重构算法中凸松弛法在恢复效率方面、贪婪追踪法在恢复精度方面存在不足,基于遗传算法迭代优化的思想,结合模拟退火以及多种群算法的优势,提出了基于模拟退火遗传算法和基于多种群遗传算法的启发式稀疏重构算法。所提算法均从传统遗传算法易陷入局部最优解的缺陷出发,分别通过保持个体间的差异性和提高种群多样性来搜索待求稀疏信号的全局最优解,并通过理论分析证明了所提算法参数选取及搜索策略的有效性。此外,以阵列信号处理中空间信源的波达方向（DOA）估计问题为例,验证所提算法的有效性。仿真结果表明,相较于正交匹配追踪OMP算法和基于l1范数奇异值分解的l1-SVD算法,所提算法提高了DOA估计的精度,且降低了运算复杂度,使其快速收敛至全局最优解。     

A hybrid shuffled complex evolution approach with pattern search for unconstrained optimization

The difficulties associated with using classical mathematical programming methods on complex optimization problems have contributed to the development of alternative and efficient numerical approaches. Recently, to overcome the limitations of classical optimization methods, researchers have proposed a wide variety of meta-heuristics for searching near-optimum solutions to problems. Among the existing meta-heuristic algorithms, a relatively new optimization paradigm is the Shuffled Complex Evolution at the University of Arizona (SCE-UA) which is a global optimization strategy that combines concepts of the competition evolution theory, downhill simplex procedure of Nelder-Mead, controlled random search and complex shuffling. In an attempt to reduce processing time and improve the quality of solutions, particularly to avoid being trapped in local optima, in this paper is proposed a hybrid SCE-UA approach. The proposed hybrid algorithm is the combination of SCE-UA (without Nelder-Mead downhill simplex procedure) and a pattern search approach, called SCE-PS, for unconstrained optimization. Pattern search methods are derivative-free, meaning that they do not use explicit or approximate derivatives. Moreover, pattern search algorithms are direct search methods well suitable for the global optimization of highly nonlinear, multiparameter, and multimodal objective functions. The proposed SCE-PS method is tested with six benchmark optimization problems. Simulation results show that the proposed SCE-PS improves the searching performance when compared with the classical SCE-UA and a genetic algorithm with floating-point representation for all the tested problems. As evidenced by the performance indices based on the mean performance of objective function in 30 runs and mean of computational time, the SCE-PS algorithm has demonstrated to be effective and efficient at locating best-practice optimal solutions for unconstrained optimization.     

A novel two-phase Hilbert-scan-based search algorithm for block motion estimation using CTF data structure

Motion estimation is one of the kernel issues in MPEG series. In this correspondence, a novel two-phase Hilbert-scan-based search algorithm for block motion estimation is presented. First in the intra-phase, a segmentation of the Hilbert curve is applied to the current block, then a novel coarse-to-fine data structure is developed to eliminate the impossible reference blocks in the search window of the reference frame. In the inter-phase, a new prediction scheme for estimating the initial motion vector of the current block is presented. Experimental results reveal that when compared to the GAPD algorithm, our proposed algorithm has better execution time and estimation accuracy performance. Under the same estimation accuracy, our proposed algorithm has better execution time performance when compared to the FS algorithm. In addition, when comparing with the TSS algorithm, our proposed algorithm has better estimation accuracy performance, but has worse execution time performance.     

Block matching algorithm for motion estimation based on Artificial Bee Colony (ABC)

Block matching (BM) motion estimation plays a very important role in video coding. In a BM approach, image frames in a video sequence are divided into blocks. For each block in the current frame, the best matching block is identified inside a region of the previous frame, aiming to minimize the sum of absolute differences (SAD). Unfortunately, the SAD evaluation is computationally expensive and represents the most consuming operation in the BM process. Therefore, BM motion estimation can be approached as an optimization problem, where the goal is to find the best matching block within a search space. The simplest available BM method is the full search algorithm (FSA) which finds the most accurate motion vector through an exhaustive computation of SAD values for all elements of the search window. Recently, several fast BM algorithms have been proposed to reduce the number of SAD operations by calculating only a fixed subset of search locations at the price of poor accuracy. In this paper, a new algorithm based on Artificial Bee Colony (ABC) optimization is proposed to reduce the number of search locations in the BM process. In our algorithm, the computation of search locations is drastically reduced by considering a fitness calculation strategy which indicates when it is feasible to calculate or only estimate new search locations. Since the proposed algorithm does not consider any fixed search pattern or any other movement assumption as most of other BM approaches do, a high probability for finding the true minimum (accurate motion vector) is expected. Conducted simulations show that the proposed method achieves the best balance over other fast BM algorithms, in terms of both estimation accuracy and computational cost.     

Block-matching algorithm based on harmony search optimization for motion estimation

Motion estimation is one of the major problems in developing video coding applications. Among all motion estimation approaches, Block-matching (BM) algorithms are the most popular methods due to their effectiveness and simplicity for both software and hardware implementations. A BM approach assumes that the movement of pixels within a defined region of the current frame can be modeled as a translation of pixels contained in the previous frame. In this procedure, the motion vector is obtained by minimizing a certain matching metric that is produced for the current frame over a determined search window from the previous frame. Unfortunately, the evaluation of such matching measurement is computationally expensive and represents the most consuming operation in the BM process. Therefore, BM motion estimation can be viewed as an optimization problem whose goal is to find the best-matching block within a search space. The simplest available BM method is the Full Search Algorithm (FSA) which finds the most accurate motion vector through an exhaustive computation of all the elements of the search space. Recently, several fast BM algorithms have been proposed to reduce the search positions by calculating only a fixed subset of motion vectors despite lowering its accuracy. On the other hand, the Harmony Search (HS) algorithm is a population-based optimization method that is inspired by the music improvisation process in which a musician searches for harmony and continues to polish the pitches to obtain a better harmony. In this paper, a new BM algorithm that combines HS with a fitness approximation model is proposed. The approach uses motion vectors belonging to the search window as potential solutions. A fitness function evaluates the matching quality of each motion vector candidate. In order to save computational time, the approach incorporates a fitness calculation strategy to decide which motion vectors can be only estimated or actually evaluated. Guided by the values of such fitness calculation strategy, the set of motion vectors is evolved through HS operators until the best possible motion vector is identified. The proposed method has been compared to other BM algorithms in terms of velocity and coding quality. Experimental results demonstrate that the proposed algorithm exhibits the best balance between coding efficiency and computational complexity.     

Fast simulated annealing hybridized with quenching for solving job shop scheduling problem

Various heuristic based methods are available in literature for optimally solving job shop scheduling problems (JSSP). In this research work a novel approach is proposed which hybridizes fast simulated annealing (FSA) with quenching. The proposed algorithm uses FSA for global search and quenching for localized search in neighborhood of current solution, while tabu list is used to restrict search from revisiting previously explored solutions. FSA is started with a relatively higher temperature and as search progresses temperature is gradually reduced to a value close to zero. The overall best solution (BS) is maintained throughout execution of the algorithm. If no improvement is observed in BS for certain number of iterations then quenching cycle is invoked. During quenching cycle current temperature is reduced to nearly freezing point and iterations are increased by many folds, as a result of this change search becomes nearly greedy. The strength of the proposed algorithm is that even in quenching mode escape from local optima is possible due to use of Cauchy probability distribution and non-zero temperature. At the completion of quenching cycle previous values of search parameters are restored and FSA takes over, which moves search into another region of solution space. Effectiveness of proposed algorithm is established by solving 88 well known benchmark problems taken form published work. The proposed algorithm was able to solve 45 problems optimally to their respective best known values in reasonable time. The proposed algorithm has been compared with 18 other published works. The experimental results show that the proposed algorithm is efficient in finding solution to JSSP.     

Truss optimization on shape and sizing with frequency constraints based on parallel genetic algorithm

Truss shape and sizing optimization under frequency constraints is extremely useful when improving the dynamic performance of structures. However, coupling of two different types of design variables, nodal coordinates and cross-sectional areas, often lead to slow convergence or even divergence. Because shape and sizing variables coupled increase the number of design variables and the changes of shape and sizing variables are of widely different orders of magnitude. Otherwise, multiple frequency constraints often cause difficult dynamic sensitivity analysis. Thus optimal criteria and mathematical programming methods have considerable limitations on solving the problems because of needing complex dynamic sensitivity analysis and being easily trapped into the local optima. Genetic Algorithms (GAs) show great potentials to solve the truss shape and sizing optimization problems. Since GAs adopt global probabilistic population search techniques and require no gradient information. The improved genetic algorithms can effectively increase the solution quality. However, the serial GA is computationally expensive and is limited on gaining higher quality solutions. To solve the truss shape and sizing optimization problems with frequency constraints more effectively and efficiently, a Niche Hybrid Parallel Genetic Algorithm (NHPGA) is proposed to significantly reduce the computational cost and to further improve solution quality. The NHPGA is to blend the advantages of parallel computing, simplex search and genetic algorithm with niche technique. Several typical truss optimization examples demonstrate that NHPGA can significantly reduce computing time and attain higher quality solutions. It also suggests that the NHPGA provide a potential algorithm architecture, which effectively combines the robust and global search characteristics of genetic algorithm, strong exploitation ability of simplex search and computational speedup property of parallel computing.