共查询到20条相似文献,搜索用时 62 毫秒
1.
主动轮廓模型主要用于计算机视觉与图像处理,尤其是用于检测对象的边界,是目前流行的图像分割算法,其主要优点是无论图像的质量如何,总可以抽取得到光滑、封闭的边界.通过使用梯度矢量流(Gradient Vector Flow,GVF)主动轮廓模型对设备内壁图像进行分割,克服了传统主动轮廓模型不能检测目标凹陷处,初始轮廓必须靠近物体边缘,以及收敛速度较慢的缺点.实验结果表明该算法不仅可以获得较好的设备内壁图像分割效果,而且能提高轮廓提取的速度,为标定设备内壁有损部位的几何参数提供良好的环境. 相似文献
2.
《现代电子技术》2016,(12):99-102
针对红外视频的移动目标检测算法在还原目标时,目标的轮廓还原准确率较低的问题,提出一种基于轮廓与背景消除的红外视频移动目标检测方案。首先,从视频中选取一些不含目标对象的帧,对选取的帧进行统计处理并建立背景帧;分析背景的场景变化建立变化的自适应背景帧,并将背景帧的杂波过滤掉;然后,使用Canny边缘检测和K-means聚类检测目标轮廓并将目标轮廓从背景提取出来;使用形态学的边缘连通算法将目标轮廓进行关闭与Flood-fill填充处理获得目标对象的形状。对比实验结果证明,相较于其他红外视频移动目标检测算法,该算法获得了较好的目标轮廓与形状,同时,该算法的检测率与虚警率性能以及每帧的处理时间均较优。 相似文献
3.
4.
5.
利用模糊特征改进Snakes模型的图像分割 总被引:1,自引:1,他引:0
利用模糊特征自适应地控制曲线法向力场改进参数主动轮廓模型,改进后的模型可以对弱边缘、无边缘区域和纹理图像进行分割。曲线法向力场加速了曲线收敛到目标区域边界,改进了抓取范围和提取凹区域的能力。对弱边缘图像、医学图像和纹理的分割实验表明,新方法具有良好的性能。 相似文献
6.
7.
提出了一种图像引力和Mumford2Shah (M - S)模型的快速曲线演化算法,该算法分两
步实现:首先利用图像引力,对参数主动轮廓线模型中的控制点进行寻优,使控制点能快速地收敛到图像的边缘附近;然后利用插值算法,得到目标较粗糙的大致轮廓,再使用M - S模型和窄带方法得到准确的轮廓。该算法避免了参数主动轮廓线模型不能收敛到物体凹陷的边缘和对初始位置敏感等问题,也避免了M - S模型需要对所有图像数据进行计算,计算量大等问题,实现了对参数和几何主动轮廓线模型进行了有效结合。 相似文献
8.
图像边缘提取的区域联合分割与主动轮廓模型 总被引:1,自引:0,他引:1
在目标的识别与跟踪处理中,目标图像的边缘提取是一项关键技术。采用边缘区域分割和主动轮廓C-V模型算法,而C-V模型更适用于水下的球体、椭球体边缘检测,具有提取的边缘连续的优点。当然,在处理不同的实际问题时,针对环境条件和要求的不同,可以选择适合的算子进行图像边缘提取。 相似文献
9.
传统二维图像轮廓识别算法通常是在图像中找到边缘,根据设定的条件将边缘像素组合起来形成轮廓。针对边缘检测算法通常需要对不同的图像设定不同的阈值,对不同类型的图像很难找出统一的阈值的问题,文中提出了一种同时利用图像中颜色与线段信息的彩色图像轮廓提取算法,算法采用自顶向下的颜色空间融合和自底向上的线段检测的方法,在初步获取边缘信息之后,综合利用检测结果生成目标的轮廓。算法的优点在于不需要进行阈值调整,实验表明:该方法可以有效地提取彩色图像中的目标轮廓。 相似文献
10.
传统Snake模型存在着对轮廓的初始化敏感,对高噪声图像易陷入局部极小值,以及对具有狭长深度凹陷区域的图像无法获得正确轮廓等问题.本文提出了一种基于边缘与区域信息的主动轮廓模型R-Snake(Region Snake).该模型通过文中设计的图像变换算子,并结合区域积分与曲线积分间转化的Green公式,导出了包含目标图像区域信息的区域力.然后由力平衡方程将该区域信息自然直接地引入到主动轮廓提取模型中,从而实现图像的轮廓提取.由于该模型同时利用了图像的区域信息和梯度信息来引导轮廓曲线的演化,使得本文方法不仅扩大了轮廓初始化的范围,降低了对图像噪声的敏感性,而且还增加了轮廓曲线收敛到真实边界的能力.实验结果表明,本文方法具有很强的适应性和鲁棒性,尤其是对高噪声图像和具有狭长深度凹陷的图像获得了优于传统Snake模型的结果. 相似文献
11.
12.
13.
Objects within a binary image are efficiently encoded by traversing their outlines to produce a string of codes each of which defines a directed line to the next point on the outline. Most commonly four or eight codes are defined because these are all that are necessary to define a move from one pixel to an adjacent pixel on a square grid. Many features of the binary objects or blobs may be extracted directly from the code string which describes the outline or contour. The paper draws attention to the fundamental nature of the crack coding scheme, shows that the chain code and midcrack code may be derived from the crack code, and proves relationships between simple measures of perimeter and area as obtained from the different codes and from bit quads. An enclosing octagon is introduced 相似文献
14.
A discrete dynamic contour model 总被引:25,自引:0,他引:25
A discrete dynamic model for defining contours in 2-D images is developed. The structure of this model is a set of connected vertices. With a minimum of interaction, an initial contour model can be defined, which is then automatically modified by an energy minimizing process. The internal energy of the model depends on local contour curvature, while the external energy is derived from image features. Solutions are presented to avoid undesirable deformation effects, like shrinking and vertex clustering, which are common in existing active contour models. The deformation process stops when a local minimum of the energy function is reached. The final shape of the model is a reproducible approximation of the desired contour. Results of applying the method to computer-generated images, as well as clinical images, are presented. 相似文献
15.
16.
本文以生物视觉功能及其机理有关的研究成果为主要依据,通过分析、建模和模拟计算等方式建立进化的图像目标轮廓组织整合的理论、方法和模型。首先通过分析初级视觉通道内(LGN—〉视觉皮层V1区—〉视觉皮层V2区神经元)目标轮廓的组织整合机理,建立具有广泛视觉心理学和生理学基础人工神经网络模型;然后将该模型应用于自然场景图像的目标轮廓组织及整合的模拟计算中,并根据实验结果对模型做出分析与评估。 相似文献
17.
人脸外轮廓线的提取方法 总被引:3,自引:1,他引:2
传统的边缘提取算子提取人脸的外轮廓线很难达到理想的效果。文中提出一种基于人脸的几何特征的外轮廓线提取方法:首先采用结合SOBEL和CANNY算子提取边缘信息,然后剂用人脸外轮廓特点去掉杂散的冗余边缘和边缘修补。该算法能够较好地提取人脸的轮廓线,而且定位精度高,满足人脸轮廓特征提取的要求。采用本方法对50多幅人脸图像进行实验,结果验证了方法的有效性。 相似文献
18.
Conformal Snake algorithm for contour detection 总被引:1,自引:0,他引:1
A novel and effective modification of the original Snake algorithm is proposed. The modification can improve the capability of the algorithm to detect boundaries with sharp comers or concave parts without the need to introduce external forces. The essential idea is to apply conformal mapping to transform the image so that the object boundary in the new domain can be captured by the Snake algorithm 相似文献
19.
A neural network-based stochastic active contour model (NNS-SNAKE)for contour finding of distinct features 总被引:2,自引:0,他引:2
Contour finding of distinct features in 2-D/3-D images is essential for image analysis and computer vision. To overcome the potential problems associated with existing contour finding algorithms, we propose a framework, called the neural network-based stochastic active contour model (NNS-SNAKE), which integrates a neural network classifier for systematic knowledge building, an active contour model (also known as the "Snake") for automated contour finding using energy functions, and the Gibbs sampler to help the snake to find the most probable contour using a stochastic decision mechanism. Successful application of the NNS-SNAKE to extraction of several types of contours on magnetic resonance (MR) images is presented. 相似文献
20.
Souvaine D.L. Bjorling-Sachs I. 《Proceedings of the IEEE. Institute of Electrical and Electronics Engineers》1992,80(9):1449-1470
The problem of finding the contour of the union of a collection of polygons in which all vertices have integer coordinates and the slopes of the sides belong to a finite fixed collection of orientations is studied from two perspectives. The first is that of determining which algorithms for finding the contour of the union for rectilinear polygons and/or rectangles can be generalized to handle polygons with sides of two or more additional directions while still retaining efficiency. The second is that of determining which algorithms for computing the union of arbitrary polygons in general position can be revised to handle degeneracies and possibly to take advantage of the restricted number of orientations to improve robustness, efficiency, or both. Three distinct rectilinear algorithms are detailed, and general preprocessing and postprocessing procedures which allow all three rectilinear algorithms to operate on polygons of additional orientations are presented. Two general algorithms for computing the contour of union based on line segment intersection algorithms and plane sweeping are also presented and analyzed 相似文献