改进的十字菱形搜索算法INCDS
- 格式:pdf
- 大小:291.93 KB
- 文档页数:5
毕业论文(设计) 2013届通信工程专业 0913071 班级题目菱形搜索运动估计算法研究及实现姓名学号指导教师职称二零一三年五月二十四日内容提要运动估计是视频压缩编码中的核心技术之一。
采用运动估计和运动补偿技术可以消除视频信号的时间冗余,从而提高编码效率;运动估计搜索算法是帧间编码的基础,常用的运动估计搜索算法采用在搜索区域内搜索最佳绝对误差和(SAD,Sum of Absolute Differences)匹配点来进行宏块匹配,获得宏块的运动矢量。
不同的搜索方法在搜索最佳SAD点上采用不同的搜索策略。
常见的快速搜索算法有三步法、新三步法、四步法、块梯度下降法以及菱形搜索算法等,本文主要研究菱形搜索运动估计算法并实现,首先阐述了课题的背景与意义和运动估计的研究现状,其次详细介绍了运动估计的原理以及典型块运动估计算法,分析它们的技术特点,然后重点介绍了菱形搜索算法,并在Visual C++ 6.0环境下编写程序代码将之实现,最后进行仿真得出实验结果。
关键词视频压缩;运动估计;块匹配;菱形搜索The Realization Of Diamond Searching MotionEstimation AlgorithmAuthor: Tutor:AbstractMotion estimation is the video compression coding technology of the core. Using motion estimation and motion compensation techniques can eliminate temporal redundancy of the video signal, thereby improving the encoding efficiency; motion estimation search algorithm based on inter-coded, the common motion estimation search algorithm of the search area to search the best absolute error and SAD (Sum of Absolute Differences) matching points to the macro block matching motion vector of the macro block obtained. Different search method searches for the best SAD point different search strategies. Common fast search algorithm has three steps, the new three-step method, four-step, block gradient descent and diamond search algorithm, etc. This paper studies the diamond search motion estimation algorithm and implementation, first describes the background and significance of the subject and motion estimation research status, followed by details of the motion estimation principle and the typical block motion estimation algorithm to analyze their technical characteristics, and then focuses on the diamond search algorithm, and the Visual C + + 6.0 environment to prepare the program code of the implementation, and finally the simulation The experimental results obtained.KeywordsVideo Compression Motion Estimation Block Matching Diamond Search目录内容提要................................. 错误!未定义书签。
基于 Harris的改进SIFT算法卢明;张晶【摘要】SIFT algorithm is a feature extraction algorithm with scale invariance,which can effec-tively extract local features of images.In order to improve the real time while improving its ro-bustness,SIFT algorithm is improved by using the local autocorrelation function of the first-order partial derivatives of the image gray from Harris algorithm instead of Hessian matrix to eliminate border effects.The improved algorithm can effectively remove the feature points in boundary posi-tion or position which is not bright enough when using Hessian matrix to eliminate border effects in SIFT algorithm,and it can improve the quality of image stitching.Experiments show that the improved algorithm can improve the detection time and robustness of SIFT algorithm,and it a-chieves good results in the production process of virtual roaming system.%SIFT算法是一种具有尺度不变性的特征提取算法,能够有效地提取图像中的局部特征。
基于高斯金字塔的图像运动估计算法王斌;何中市;伍星;贾媛媛【摘要】运动估计是图像超分辨率重建中的关键环节,直接影响超分辨重建的结果质量。
为减少运动估计搜索点数,提高搜索速度,提出一种基于高斯金字塔分层思想的小十字形搜索算法。
算法把图像构建成一个两层高斯金字塔,在上层使用小十字形搜索算法估计初始运动矢量,并通过提前终止策略来提前结束搜索;在下层以上层估计结果作为初始值,估计最终图像运动矢量。
在标准图像序列上进行的实验结果表明,该算法在保持搜索精度的前提下能明显提高搜索速度;特别对于在运动偏差较大的情况下,提高效果更显著。
%Motion estimation is an essential process in the super-resolution reconstruction which directly effects the quality of resultant image. To reduce the searching points of motion estimation and improve the search speed, this paper proposes an improved small cross-diamond search algorithm based on Gaussian Pyramid. A Gaussian Pyramid with two layers is constructed. The first layer is used for estimating the initial motion vector by using the small cross-diamond search algo-rithm, and the early termination strategy is adopted to terminate the algorithm in advance, while the second layer evalu-ates the final offset value based on the initial motion vector produced in the first layer. Experiment is conducted with stan-dard image sequences. The experimental results show that the proposed algorithm can greatly improve search speed within the desirable search accuracy, and especially for the sequences with large motion deviation.【期刊名称】《计算机工程与应用》【年(卷),期】2015(000)007【总页数】5页(P174-178)【关键词】高斯金字塔;小十字形搜索算法;提前终止策略;图像运动矢量【作者】王斌;何中市;伍星;贾媛媛【作者单位】重庆大学计算机学院,重庆 400044;重庆大学计算机学院,重庆400044;重庆大学计算机学院,重庆 400044;重庆大学计算机学院,重庆400044【正文语种】中文【中图分类】TP3091 引言在大部分图像数据应用领域里,为了获取高质量图片,常常要求获得具有高分辨率的图像,然而现实世界中,能够获取的多为低分辨率图像。
—227—一种用于运动估计的十字六边形搜索算法吕 瑞,卿粼波,何小海,龙建忠(四川大学电子信息学院图像信息研究所,成都 610064)摘 要:提出了一种十字六边形搜索算法用于快速运动估计。
该算法利用了运动矢量的中心偏置性和相关性,运动估计时通过预测确定搜索起始点,在搜索前期利用十字模板结合提前退出技术优先搜索起始点附近的局部区域,后期则改用六边形模板扩大搜索范围并完成运动估计。
实验证明该算法与原始的六边形搜索算法相比平均减少了45%的搜索点数,与一些新的快速搜索算法相比,在搜索精度基本相似的情况下也有效地降低了运动估计的运算复杂度。
关键词:运动估计;提前退出;十字模板;六边形搜索Cross Hexagon Search Algorithm for Motion EstimationLV Rui, QING Linbo, HE Xiaohai, LONG Jianzhong(Institute of Image Imformation, College of Electronics and Information Engineering, Sichuan University, Chengdu 610064)【Abstract 】A cross hexagon search(CHS) algorithm is proposed for fast motion estimation(ME). CHS algorithm utilizes both the center-based distribution characteristic of motion vectors(MVs) and the correlation among MVs. In ME, the initial search points are decided by prediction at first,then at early searching stage CHS employs cross search pattern and halfway-stop technique in searching the local areas around the initial search points priority, and at later stage CHS enlarges searching range and completes ME by using hexagon search patterns. Experiment results show that CHS gains 45% average reduction in average searching point(ASP) over original hexagon-based search(HEXBS) algorithm, and that CHS can effectively reduce the computational complexity of ME over some new fast searching algorithms with similar searching precision. 【Key words 】motion estimation; halfway-stop; cross searching pattern; hexagon-based search (HEXBS)计 算 机 工 程Computer Engineering 第33卷 第13期Vol.33 No.13 2007年7月July 2007·多媒体技术及应用·文章编号:1000—3428(2007)13—0227—03文献标识码:A中图分类号:TP391在视频压缩编码中,运动估计是减少视频序列时间冗余度的有效手段,其运算效率对整个编码系统的性能有着重大影响。