A Road Extraction Method by an Active Contour Model with Inertia and Differential Features
A Road Extraction Method by an Active Contour Model with Inertia and Differential Features
复制标题
DOI：
10.1093/ietisy/e89-d.7.2257
复制
发表时间：
2006-07
期刊：
影响因子：
--
通讯作者：
Hiroaki Sawano;M. Okada
中科院分区：
文献类型：
--
作者：
Hiroaki Sawano;M. Okada
在本文中，我们提出了一种基于活动轮廓模型（蛇）的道路对象提取技术，考虑到惯性和电影中的差异功能。不同的能量函数可以应用于snake，以便使用各种对象和各种环境的信息。使用许多方法来跟踪移动对象，snake可以逐帧应用于场景。帧中的控制点的初始位置可以参考前一帧中的结果。我们专注于惯性，在以前和现在的帧中的对象形状之间的工作。在这项研究中，惯性是控制点抵抗其在图像空间中运动状态变化的趋势。我们引入了一个外部的能量蛇的控制点的惯性的基础上。引入基于道路几何微分特征的内能函数，实现了直线、圆弧和S形道路段的平滑提取。所提出的方法被应用到提取道路几何形状从一个电影由一个摄像机配备在flont的车辆。实验结果表明，该方法能够平滑地提取道路几何特征，提高了道路几何特征提取的鲁棒性。
In this paper we propose a road object extraction technique based on an active contour model (snake) considering inertia and differential features in a movie. Different energy functions can be applicable to snake in order to use information of various objects and various environments. Using many methods for tracking a moving object, snake can be applied to a scene frame by frame. Initial positions of the control points in a frame can refer to the results in the previous frame. We focus on the inertia that works between object shapes in the previous and present frames. In this research inertia is the tendency of a control point to resist its changes in its state of motion in an image space. We introduce an external energy for snake based on inertia of control points. Internal energy functions based on differential features of road geometry are also introduced to extract straight, circular and S-shaped road segments smoothly. The proposed method is applied to extract road geometry from a movie taken by a camera equipped on the flont of a vehicle. Experimental results indicate the availability of the proposed method which is to extract road geometry smoothly and to improve its robustness.