欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了满足移动机器人室内实时导航的需要,在单目视觉障碍检测中引入了LVQ神经网络分类器,在HSV空间内分割图像,有效地消除了环境因素对障碍检测的影响;然后从摄像机模型几何关系出发,推导出图像坐标与机器人坐标的转换关系,实现对障碍物的快速准确定位;该方法具有准确度高和实时性强的优点,并且对环境光线变化和阴影有较强的自适应能力,实际环境中的实验结果证明了该方法的有效性。
英文摘要：
A LVQ neural network classifier is introduced into monocular vision-based obstacle detection to segment image in HSI color space to meet the demand of real-time indoor navigation of mobile robot.Effect of environmental factor is effectively removed.Then the transformational relation between image coordinate and robot coordinate is deduced from camera model geometric relationship to implement rapid and accurate obstacle position.The proposed method is of high accuracy and real-time.It also has the adaptability to changing light condition and shadow.Results of experiment in practical environment demonstrates the validity of the method.
同期刊论文项目
同项目期刊论文
Environmental/economic power dispatch problem using multi-objective differential evolution algorithm
Intelligent Foreign Particle Inspection Machine for Injection Liquid Examination Based on Modified P
Existence and exponential stability of almost periodic solution for shunting inhibitory cellular neu
期刊信息
位置：