线激光条纹中心亚像素定位研究

摘要

For the limitation that existing algorithms have not a good tradeoff between speed and precision,an adaptive threshold weighted barycenter method is proposed to obtain sub-pixel location of the line laser stripe,according to the red channel of an image has better texture.The method takes the red channel separated from the original image as the research object,and the laser stripe is rough extracted by Otsu adaptive threshold value algorithm.Finally,the sub-pixel coordinates of the laser stripe are obtained by weighted barycenter method.The method eliminates the influence of background in the extraction of laser stripe process,increases the weight of brighter coordinates,narrows the range of valid data and reduces the computation quantity of barycenter method.The result of the experiment shows the algorithm has faster speed and higher calculation accuracy,which can be used in the real time scanning and ranging system.%针对现有算法不能兼顾线激光条纹提取速度和精度的缺陷,根据激光条纹红色通道分量的纹理特性,提出自适应阈值加权重心法进行激光条纹的亚像素定位.该方法以含激光条纹的图像中的红色通道分量为研究对象,将Otsu算法得到的自适应阈值对激光条纹进行粗提取,最后利用加权重心法获取激光条纹的亚像素坐标.自适应阈值加权重心法消除了背景对提取激光条纹的影响,增加了光强较强点在重心法中的权重,缩小了有效数据的范围,减少了重心法的计算量.实验表明,该方法具有较快速度和较高的精度,可应用于实时扫描测距系统中.