UC知道十万火急,请教高手啊!!!!!!!!!!!!!!!!!!
来源：百度知道 编辑：UC知道 时间：2024/05/22 11:54:13
Particular advantages of the GVF snake over a traditional
snake are its insensitivity to initialization and its ability to
move into boundary concavities. As we show in this paper,
its initializations can be inside, outside, or across the object’s
boundary. Unlike pressure forces, the GVF snake does not
need prior knowledge about whether to shrink or expand
toward the boundary. The GVF snake also has a large capture
range, which means that, barring interference from other
objects, it can be initialized far away from the boundary. This
increased capture range is achieved through a diffusion process
that does not blur the edges themselves, so multiresolution
1057–7149/98$10.00 ã 1998 IEEE
360 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 7, NO. 3, MARCH 1998
methods are not needed. The external force model that is
closest in spirit to GVF is the distance potential forces of
Cohen and Cohen [12]. Like GVF, the
snake are its insensitivity to initialization and its ability to
move into boundary concavities. As we show in this paper,
its initializations can be inside, outside, or across the object’s
boundary. Unlike pressure forces, the GVF snake does not
need prior knowledge about whether to shrink or expand
toward the boundary. The GVF snake also has a large capture
range, which means that, barring interference from other
objects, it can be initialized far away from the boundary. This
increased capture range is achieved through a diffusion process
that does not blur the edges themselves, so multiresolution
1057–7149/98$10.00 ã 1998 IEEE
360 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 7, NO. 3, MARCH 1998
methods are not needed. The external force model that is
closest in spirit to GVF is the distance potential forces of
Cohen and Cohen [12]. Like GVF, the
特别是优势的gvf蛇超过传统
蛇是其敏感的初始化和能力
搬进边界concavities 。正如我们表明，在这方面的文件，
其初始化可以内，外，或横跨对象的
边界。不同势力的压力， gvf蛇不
需要事先了解是否缩小或扩大
对边界。该gvf蛇也有一个大捕捉
范围，这意味着，除非有其他的干扰
物体，它可以初始化远离边界。这个
增加捕捉范围是取得通过扩散过程
这并不模糊的边缘，因此，多分辨率
1057-7149/98 $ 10.00 ã 1998年的IEEE
360的IEEE交易的图像处理，第一卷。 7 ，没有。三， 1998年3月
方法是没有必要。对外力模型是
在最接近的精神gvf是距离的潜在力量
科恩和科恩[ 12 ] 。像gvf ，这些力量源于
1边缘图的形象和可以提供大量捕获
范围。我们所显示的，不过，不像gvf ，距离潜在
部队不能移动蛇到边界concavities 。我们
相信这是一个财产的所有保守势力认为，
的特点，几乎所有的蛇外部势力，并探索
非保守外部势力，例如gvf ，是一个重要的
方向，为今后的研究在主动轮廓模型。
我们注意到，部分工作报告，在这份文件
出现在会议的文件[ 18 ] 。
二。背景
答：参数Snake模型
传统的蛇是一个曲线，
，移动，通过空间域的图像
尽量减少能源的功能
（ 1 ）
在哪里和是加权参数控制
蛇的紧张和僵硬，分别和和
是指第一和第二的衍生物与尊重
。外部能量函数是来自形象
因此，它需要就其规模较小的价值的特点，兴趣，
如界限。由于灰度级图像，看
作为一个函数的连续变量的立场，典型
对外精力设计带领一个主动轮廓线走向
一步的优势[ 1 ]
（ 2 ）
（ 3 ）
那里是一个两维高斯函数与
标准偏差，是梯度算子。如果
形象是一个线描（黑与白） ，然后适当