节点文献
自然环境下果实定位的误差修正方法研究
【摘要】 对用双目立体视觉技术来测定番茄果实空间位置中存在的误差原因进行了分析,提出采用神经网络的方法对误差进行修正,建立网络模型和训练系统,并对测试样本进行仿真实验。实验结果表明,绝对误差可以控制在0~5mm之间,达到了收获机器人定位精度的要求。
【Abstract】 The causes of errors existing in the measurement of threedimensional coordinates of the tomato with the help of binocular vision technology are analyzed.Then it is pointed out that the errors can be amended by means of building a neural net model and a corresponding training system. A simulation experiment is also carried out on the test sample. The experiment indicates that the errors can be controlled in the range of 0 -5 mm, thus meeting the requirement for positioning of robots harvesting fruits.
【关键词】 双目立体视觉;
空间定位;
误差分析;
神经网络;
误差修正;
【Key words】 Binocular Three-dimensional Vision Three-dimension Orientation Error Analysis Neural Net Error Amendment;
【Key words】 Binocular Three-dimensional Vision Three-dimension Orientation Error Analysis Neural Net Error Amendment;
【基金】 江苏省自然科学基金资助项目(编号:BK2003046)
- 【文献出处】 机械制造 ,Machinery , 编辑部邮箱 ,2006年06期
- 【分类号】TP242
- 【被引频次】2
- 【下载频次】125