摘要
为了更好地解决智能激光切割机工件识别与定位问题,提出了双相机定位方法,全局相机负责对象识别与粗定位,局部相机负责精定位,通过迭代逼近方法削除局部相机镜头畸变对测量精度的影响。改进的圆投影匹配方法利用半径表快速计算圆投影向量,由粗到精的匹配策略和极坐标变换旋转角度计算,明显地提高了匹配运算速度。实验证明本识别算法和定位方法实用且精度高。
In order to solve workpiece recognition and positioning problem of intelligent laser cutting machine, a dual-camera positioning method is proposed, global camera processes object recognition and coarse positioning, local camera processes fine positioning, eliminates the effect from lens distortion of local camera on measurement precision by iterative approximation method. Improved circular projection matching algorithm could fast calculate circular projection vector using radius table, adopted coarse -to -fine matching strategy and could compute rotation angle by polar coordinates transformation, obviously improved the matching speed. Experiments proved the matching algorithm and positioning method are practical and have high accuracy.
出处
《微计算机信息》
2011年第4期77-78,82,共3页
Control & Automation
关键词
智能激光切割机
双相机定位
圆投影匹配
intelligent laser cutting machine
dual-camera positioning
circular projection matching
作者简介
硕士研究生
