首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到18条相似文献,搜索用时 125 毫秒
1.
为了更好地分割图像,对传统Split-merge算法作出改进:PCNN先进行分裂,用一种简化的Mumford-Shah模型进行合并,使得分裂阶段不仅无效分割减少,而且无方块效应,对边缘定位准确;合并阶段能够理想地将分裂后的区域合并为感兴趣的前景与不感兴趣的背景,误合并与欠合并大幅减小.对Papav, Monkey, Twoman图像的仿真表明,其分割结果和运行时间均优于AMS,MBMS算法.仿真结果证明,本文算法是一种适用于图像分割的有效算法.  相似文献   

2.
基于边缘检测与分裂合并的图像分割算法   总被引:2,自引:0,他引:2  
针对传统分裂合并算法容易产生方块效应与过分割的缺点,提出了一种结合边缘检测和分裂合并的图像分割算法.该算法直接利用图像的边缘信息进行分裂,不断将图像分裂为一些不规则形状的一致性区域,然后根据一定规则将相似的区域合并.实验表明,该算法能大幅减少分裂次数,并有效克服方块效应和过分割等缺点,图像分割效果较好.  相似文献   

3.
传统模糊ISODATA(Fuzzy ISODATA,FISODATA)算法中,分裂-合并操作需人工选取阈值参数。而不适当的阈值往往使算法陷入局部极值,因而得到错误的类属数并最终影响图像分割结果。为此,在模糊集理论基础上提出一种改进的自适应FISODATA算法。该算法设计了自适应分裂-合并操作,即在每次分裂-合并后,根据该次计算结果改变参数阈值,解决了人为选取参数带来的诸多问题。利用该算法对模拟图像和真实IKONOS图像进行分割实验,均能得到良好的分割结果。  相似文献   

4.
针对现有的道路标线分割方法在阴影、光照、裂缝等干扰下存在分割精度不足的问题,提出了基于分裂合并算法的道路标线分割方法.该方法通过分析道路标线图像的灰度、边缘、纹理特征,首先采用Bresenham算法进行道路标线有无判定,并根据道路标线边缘实现路面检测图像标线区域和非标线区域的分裂,然后采用基于重心连线灰度值的区域合并和基于灰度与直线边缘特征的区域合并两种方法进行道路标线区域合并,最终完成道路标线的准确分割.实验表明,该方法分割精度较高,具有良好的抗噪性能,能够很好地满足道路检测车采集图像实时处理的需求.  相似文献   

5.
分裂合并算法是一种基于区域的串行图像分割算法.在图像的分裂阶段,从图像区域的表示出发,引入Morton码,降低了算法的空间复杂度.同时在图像的相邻区域的合并阶段,提出了一种新的合并原则,增加了满足合并条件的相邻区域的匹配率,减少算法的迭代次数,提高了算法的执行效率.最后给出并分析了相关的实验数据,证明了算法的有效性.  相似文献   

6.
提出一种融合快速全局K-means与区域合并的图像分割方法。该方法利用中值滤波方法对图像去噪;运用快速全局K-means算法对图像的颜色空间进行聚类分析;结合区域合并准则,对初始分割合并得到最终的分割结果。实验表明,与同类算法比较,该方法的分割结果在图像细节方面能够很好地满足人的主观视觉。  相似文献   

7.
Split-Merge算法是图像分割领域的经典算法,思路简洁、运算效率高,但因分裂位置和形状固定,导致出现方块效应和边界定位不准确等问题.现有的各种改进算法不能从根本上解决这些问题,为此本文将基于图的分割算法纳入到Split-Merge框架中来,提出了在Split-Merge框架下进行基于图的分割.新算法把传统的基于图割值的算法作为分裂阶段的处理手段,在合并阶段采用本文提出的一种与访问次序无关的合并算法.新算法既解决了Split-Merge算法中因分裂位置和形状固定而带来的各种问题,也有效地消除了在基于图的分割中常见的"过分割"现象,取得了满意的分割效果.  相似文献   

8.
张鲲  王士同 《计算机应用》2007,27(3):673-676
提出一种新的混合多阶段无监督图像分割算法。在第一阶段,通过分水岭算法得到一幅过度分割的图像,该图像中的所有小区域作为初始聚类状态将在接下来的层次聚类阶段中被合并。在第二阶段,一种新的启发式的基于Bayesian方法和Markov随机域的计算模型被用于基于区域的层次聚类算法,该算法用来合并初始分割结果中的邻接区域,以改进分水岭算法的分割效果。深入分析了该计算模型中两个相互作用的部分。通过对多种不同种类图像使用该算法进行分割,表明这种多阶段的方法适合无监督分割,它按照视觉一致的方式合并区域,并且比传统的层次聚类算法快很多。  相似文献   

9.
基于组件树滤波及快速区域合并的分水岭分割算法   总被引:2,自引:0,他引:2  
闫沫 《计算机科学》2013,40(1):282-285
针对分水岭算法存在过分割的问题,提出一种结合组件树滤波及快速区域合并的图像分割算法。该算法在图像预处理阶段利用组件树来表示梯度图像且根据顺序极值计算分水岭的相对势能和属性,并对其进行滤波,从而减少梯度图像中的局部极小值。对滤波后的梯度图像进行分水岭初始分割,然后利用完美场景准则对初始分割结果进行快速区域合并。实验结果表明,采用组件树对梯度图像进行滤波能够减少由于噪声而产生的局部极小值,大大减少了分水岭初始分割区域数量,提高了区域合并的准确性,加快了合并速度。  相似文献   

10.
为解决传统FLICM算法需人为给定图像聚类数的问题,基于该算法通过聚类中心描述聚类的特点,设计以聚类中心为操作对象的分裂合并操作,以实现可变类图像分割.在此基础上定义分裂合并操作的接受率,不但能够有效避免算法陷入局部极值,促进其快速收敛,同时有利于参数阈值的自适应.分别利用所提出算法和传统ISODATA算法分割模拟图像和灰度纹理图像,对其结果的定性定量分析验证了所提出算法的有效性和普适性.  相似文献   

11.
基于层次与划分方法的聚类算法研究   总被引:3,自引:1,他引:3  
针对在层次聚类算法中,一个分裂或合并被执行,就不能修正,其聚类质量受到限制的缺陷,提出了利用簇间相异度及基于信息熵或整体相似度的聚类质量评价标准,在簇分裂过程中动态的进行簇的合并与分裂的算法。仿真实验结果证明,该算法具有使结果簇更紧凑和独立的效果,具有更好的聚类质量。  相似文献   

12.
Using a split and merge image segmentation algorithm on NOAA-7 AVHRR channel four data, the thermal structure of an agricultural homogeneous region (the Beauce Plateau in France) around midday is studied during 1982. It is shown that the image homogeneous areas can be identified with known natural regions: hills, forest, and the Beauce Plateau. The Beauce Plateau has a thermal structure divided into three or more areas depending on the season. Those areas are very well correlated with regions of equal soil water availability.  相似文献   

13.
Summary A simple mathematical model for analyzing the dynamics of a B-tree node is presented. From the solution of the model, it is shown that the simple technique of allowing a B-tree node to be slightly less than half full can significantly reduce the rate of split, merge and borrow operations. We call split, merge, borrow and balance operations unsafe operations in this paper. In a multi-user environment, a lower unsafe operation rate implies less blocking and higher throughput, even when tailored concurrency control algorithms (e.g., that proposed by Lehman and Yao [10]) are used. A lower unsafe operation rate also means a longer life time of an optimally initialized B-tree (e.g., compact B-tree). It is in general useful to have an analytical model which can predict the rate of unsafe operations in a dynamic data structure, not only for comparing the behavior of variations of B-trees, but also for characterizing workload for performance evaluation of different concurrency control algorithms for such data structures. The model presented in this paper represents a starting point in this direction.  相似文献   

14.
A new region based lossy compression scheme for color images is proposed. The segmentation method belongs to the split and merge category. Splitting is carried out using the watershed transform. In the merging stage, a fuzzy color preserving rule-based system and a novel one-dimensional graph structure are introduced to provide accurate results with reduced computational complexity. The compression part is based on the Shape Adaptive DCT with ΔDC correction method. The quantization matrices used have been designed according to the properties of the employed transform. Promising perceptual results for the low bit rate range compared to previously reported compression methods have also been reported.  相似文献   

15.
烟尘图像分割是图像分割方法的一个较新的应用方向。高效、准确的烟尘图像分割技术对于大气污染监控、森林火灾预警以及军事情报获取等方面都有重要的意义。将阈值分割、区域生长、区域分裂与合并等基于区域的方法用于烟尘图像分割,并对三种分割方法在烟尘图像分割中的性能表现进行了分析。研究结果对烟尘图像分割中算法的选型具有一定的参考价值。  相似文献   

16.
In this paper, we propose a new online identification approach for evolving Takagi–Sugeno (TS) fuzzy models. Here, for a TS model, a certain number of models as neighboring models are defined and then the TS model switches to one of them at each stage of evolving. We define neighboring models for an in-progress (current) TS model as its fairly evolved versions, which are different with it just in two fuzzy rules. To generate neighboring models for the current model, we apply specially designed split and merge operations. By each split operation, a fuzzy rule is replaced with two rules; while by each merge operation, two fuzzy rules combine to one rule. Among neighboring models, the one with the minimum sum of squared errors – on certain time intervals – replaces the current model.To reduce the computational load of the proposed evolving TS model, straightforward relations between outputs of neighboring models and that of current model are established. Also, to reduce the number of rules, we define and use first-order TS fuzzy models whose generated local linear models can be localized in flexible fuzzy subspaces. To demonstrate the improved performance of the proposed identification approach, the efficiency of the evolving TS model is studied in prediction of monthly sunspot number and forecast of daily electrical power consumption. The prediction and modeling results are compared with that of some important existing evolving fuzzy systems.  相似文献   

17.
一种基于模板匹配的电路照片拼接方法   总被引:6,自引:1,他引:5       下载免费PDF全文
鉴于实际工作中,经常需要将多幅图象拼成一大幅大图象的问题,提出了一种针对显微照片的新拼接方法。该方法是应用模板匹配技术来实现图象的自动拼接,由于是通过利用图象信息来选择模板,从而为模板匹配的定位创造了条件,该方法分为模板选择、模板匹配和图象拼接3步。实际运用结果表明,该方法简单易行,在电路照片的拼接中,定位较准确,且拼接效果好。  相似文献   

18.
Information about objects that split or merge is often generated dynamically as a by-product of computation or in the observation of real-world behavior. Visualization tools for such processes must not only reveal temporal patterns and spatial organization but should also accommodate on-the-fly generation of split–merge information. This paper develops a formal structure for split–merge processes that provides a suitable underpinning for dynamic visualization tools. The structure allows split–merge processes to be constructed dynamically and supports operations such as concatenation and partitioning while maintaining the underlying split–merge structure. The paper also proposes a method for abstracting split–merge substructure to assist in the visualization of large systems. The abstractions can be used to produce visual simplification and to guide layout heuristics in producing better visualizations. The implications of visualizing split–merge processes with time lines are explored. Auxiliary visualization strategies based on dynamic boxes and tree rings are introduced to enhance information content of focus areas in larger time-line visualizations. The concepts are illustrated using data from timings of a network protocol and from the detection of patterns in scientific video data.  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号