摘要
用主动视觉方法获取物体表面的空间和彩色信息需要充分利用照射在物体表面上的激光线所包含的信息,因此,准确地将激光线在图像中的位置提取出来就成为主动视觉方法中的一个关键问题。但是,由于遮挡、物体表面反射特性等的影响,所摄图像不能反映出激光线的准确和完整的信息,从而给激光线位置的准确提取带来困难。文中分析了激光带状线的提取方法和将其单像素细化的方法,并且采用三次样条函数插值方法进行补断。实验结果说明该方法能比较准确地将图像中的激光线位置提取出来。
To acquire the space and color information of object surface by active vision, we must make full use of the information contained in the laser curve illuminated on the object surface. Therefore, extracting the accurate location of laser curve in image plays a key role in active vision. Because of occlusion, reflecting characteristics of object surface and so on, the accurate and full information of laser curve of image can't be acquired directly. It makes the extracting of accurate location of laser difficult. The methods of extracting laser band and of putting it into one pixel and of making it continuous by cubic spline function interpolation method are given in this paper. Experimental results show that the accuracy of extracting the laser curve location is satisfactory.
出处
《红外与激光工程》
EI
CSCD
1998年第5期39-43,共5页
Infrared and Laser Engineering
基金
国家自然科学基金资助项目
关键词
计算机视觉
主动视觉
激光线位置
提取方法
Computer vision
Active vision
Laser
Histogram threshold method
Cubic spline function interpolation method