标记分水岭算法及区域合并的遥感图像分割

SEGMENTATION OF THE REMOTE SENSING IMAGE BASED ON METHOD OF LABELING WATERSHED ALGORITHM AND REGIONAL MERGING

陈波[1,2] 张友静[2] 陈亮[1,2]

传统的分水岭算法通常对梯度图像进行无标记分割，其结果是容易造成过度分割。本文采用了一种基于标记的分水岭算法，首先，利用Sobel边缘算子对原遥感图像进行梯度重建，获得梯度幅值图像，同时计算待分割区域的周长、面积和形态因子，并对其进行标记；然后，利用距离函数图标定种子法和等值线跟踪法获得初始分割图像；最后，利用改进的区域合并方法获得最终的分割结果。实验结果表明了该方法的有效性。[著者文摘]

Segmentation of gradient images in the traditional watershed algorithm usually has no markers, which is likely to cause excessive segmentation. This paper presents a watershed algorithm based on the label. First, the gradient images are obtained through the reconstruction of gradient by using Sobel operator and, at the same time, the perimeter, area and morphology factors of the region are computed and labeled. Then, the initial image of segmentation is acquired by using the method of distance function icon for determining the seeds and the technique of isoline tracking. Finally, the last result of segmentation is obtained by using an improved method of regional merging. The experimental result shows the effectiveness of the method.[著者文摘]

一种改进的分水岭图像分割算法

A improved watershed image segmentation algorithm

吴雷磊 王卫星

分水岭算法在医学图像分割中有着广泛的应用。针对传统分水岭分割算法对噪声敏感和严重的过分割问题．本文提出了一种基于数学形态学的改进的分水岭图像分割算法。通过前期的预处理和汇水盆地的选取，有效地抑制了过分割现象。实验表明，该算法能够对血细胞图像进行快速、准确的分割，边缘定位精确，并且能够较好的克服噪声对图像的影响。[著者文摘]

The watershed algorithm is widely used in medical image segmentation. In allusion to resolving the problems of noise sensitivity and over-segmentation existing in traditional watershed algorithm, this paper proposes an improved watershed segmentation algorithm based on morphological mathematics. Through the pretreatment in prophase and the catchment basin chosen, the phenomena of over-segmentation can be restrained. Experimental results show that the proposed algorithm can segment images of blood corpuscle quickly and accurately, and exactly locate edges. Furthermore, the influence of noisy can be averted by this way.[著者文摘]

基于分水岭和模糊C均值聚类的图像分割方法

Image segmentation based on watershed and fuzzy C-means clustering

杨洪薇 肖志涛 翁秀梅

针对分水岭算法对微弱边缘和噪声非常敏感、容易导致过分割现象的问题，提出综合运用分水岭算法和基于区域的模糊C均值聚类的图像分割方法．与单独使用分水岭方法相比，该方法不仅利用区域的灰度信息，而且考虑了区域间的空间信息．实验结果表明，本方法能有效地对图像进行分割，克服了分水岭算法的过分割问题．[著者文摘]

Watershed algorithm often causes oversegmentation because of its high sensitivity to the weak edge and the noise. To overcome its limitation, watershed algorithm was combined with region-based fuzzy C-means clustering for image segmentation. Compared with watershed algorithm, not only the grayscale information but also each region＇s spatial information is considered in this method. Experimental results show that this method can efficiently overcome the oversegmentation problem.[著者文摘]

基于分水岭变换的彩色图像分割

龚天旭 彭嘉雄

为了克服传统分水岭算法固有的过度分割问题，采用了标记-分水岭变换，提出了一种新的标记提取方法。通过计算图像中每个小区域(5×5)的均匀性而得到一幅与原图同样大小的表征原图区域均匀性的均匀性图，取均匀性图中的低谷作为标记对均匀性图进行分水岭变换就得到最终的分割结果。

一种改进的分水岭人脑MR图像分割方法

AN IMPROVED WATERSHED TRANSFORM FOR BRAIN MR IMAGE SEGMENTATION

李荣艳[1] 熊文生[2] 曹佃国[3]

针对传统分水岭算法分割图像后存在的细节不准确问题，提出了一种改进的分水岭算法．此算法将传统分水岭算法和概率相结合，通过有序队列来实现边界的计算，从而得到分水岭．实验证明，这种算法得到的结果更加准确，对人脑MR图像病灶区的分割非常重要．[著者文摘]

As to the problem of inaccurate detail in scheme segmentation, a new method of improved watershed algorithm is proposed. This method combines traditional watershed algorithm and probability, Through ordered queue, boundary is calculated and watershed is obtained. The experiment proves that the precision of this method is very high and very useful to the segmentation of brain MR scheme.[著者文摘]

基于分水岭变换和蚁群聚类的图像分割

Image segmentation method based on watersheds and ant colony clustering

杨卫莉 郭雷 赵天云 肖谷初

针对传统分水岭分割算法对噪声敏感和易于产生过分割问题，提出一种新的基于分水岭和蚁群智能聚类的图像分割方法（CWAC,Combining watersheds and ant colony clustering）。CWAC方法首先用分水岭变换对图像做初分割，然后用蚁群方法在区域之间进行聚类合并，获得最终的分割结果。CWAC不但成功地解决了分水岭存在的过分割问题，还大大提高了蚁群聚类算法的搜索效率；本文利用分水岭变换后的灰度信息和空间信息，定义了一种新的引导函数，可更准确有效引导蚁群聚类。实验结果表明CWAC可以快速准确地分割出目标，是一种有效的图像分割方法。[著者文摘]

Aimed at resolving the problems of sensitivity to noise and over-segmentation existing in traditional watershed algorithm, a new image segmentation method -CWAC is presented. First, an image is separated into a large number of small partitions by watershed algorithm and the characteristic parameters are calculated. Second, CWAC method merges different regions of homogeneity with ant colony clustering algorithm to gain result of image segmentation. CAWC algorithm can successfully solve the over-segmentation problem and at the same time it can reduce the computational times of ant colony clustering. In order to be more accurate and efficient at clustering ant colony, a new visibility based on intensity distribution and spatial information is defined. Experimental results show that CWAC can segment objective quickly and accurately and it is a practicable method for the image segmentation .[著者文摘]

基于分水岭-隐马尔可夫模型联合的语义视频对象提取技术

SEMANTIC VIDEO OBJECT EXTRACTION BY HYBRID WATERSHED-HIDDEN MARKOV RANDOM FIELD MODEL

杨树堂 汤志平

针对3C无线网络的增值服务业务，提出一种语义对象基于分水岭算法的视频帧内的分割，及基于隐马尔可夫模型的视频帧间跟踪提取技术。其主要的特点是首先采用基于标识集的分水岭算法来进行初始帧内语义视频对象的标定和分割处理，随后再进行二值化掩膜处理，最后借助隐马尔可夫测量场模型，将后续帧中视频对象的跟踪处理演化为跟踪区域与非跟踪区域的二值离散化标定问题。实验结果证明，该算法能很好地实现视频帧序列中语义视频对象的连续提取。[著者文摘]

In order to deal with the increment service of 3G wireless network, a semantic video object extracting technique based on hybrid watershed-hidden Markov random field model is put forward. Its main characteristic is that the semantic video object in the initial frame is identified and segmented by improved watershed algorithm based on identifier sets, and then the SVO is masked by two valuse 0 and 1. Finally the video object is tracked in the next frames by using the hidden Markov Random field model. The tracking process can be regarded as discrete labeling problem. The results of the experiment show that the algorithm performs well in semantic video object extraction.[著者文摘]

基于分水岭变换的目标图像的分割与计数方法

A Segment and Count Method Based on Watershed Transformation

林小竹[1] 王彦敏[2] 杜天苍[1] 田瑞卿[2]

针对在图像分析中所获得的目标物体相互粘连的情况，利用分水岭变换的模拟浸水原理，提出了一种新的、有效的分水岭分割算法。算法只需考虑图片中像素的4连通和集水盆地的边缘情况即可，使得粘连物体得到了有效的切割和分离。统计各个所得盆地即可以得到实际物体的准确个数。该分割方法成功有效地解决了目标物体粘连而造成的不便和对后续处理的干扰等问题，同时把该算法用于对粘连物体的计数方面，使本来繁琐、复杂的图像物体计数变得十分简单、便捷。是一种切实可行的图像分割与计数方法。[著者文摘]

This paper presents an efficient watershed algorithm that is based on connected components. In this algorithm, the 4-conn and edge of the catchment basin is the only thing that needs to be considered, so the algorithm is low memory consuming. Furthermore, the algorithm can make the touched objects separate from each other efficiently. It can get the exact number of the target object from counting the number of the catchment basin. It successfully solves the inconvenience brought by the touched objects and carves out an efficient way for the successive work. At same time, it is applied to count the number of objects and make the counting number of objects very easy and convenient. So, it is an efficient way to separate and account touched objects.[著者文摘]

奥运会是行情分水岭吗？

市场上关于奥运会有可能成为股市转折点的说法由来已久，进入2008年，随着奥运会日益临近，这一话题更是日益升温。“奥运不是中国股市的分水岭。”北京大学金融与证券研究中心主任曹凤岐旗帜鲜明地表达了自己的看法。他认为，奥运肯定会对中国经济产生推动作用，但并不是奥运之前股市好、奥运之后就不行了。实际上，奥运带给了中国很多更好的投资机会，奥运以后还会有更多的投资机会。“有人说在奥运之前赶快把股票都出掉，实际上肯定吃大亏，千万别出掉。”[第一段]