%0 Journal Article
%T 基于激光三角测距的锯材表面缺陷检测方法
%A 张未
%A 吕志娟
%A 徐兆军
%A 朱南峰
%J 林业工程学报
%D 2017
%R 10.13360/j.issn.2096-1359.2017.06.020
%X 为实现自动在线检测锯材表面钝棱和裂纹等缺陷,提出了一种全新的基于激光三角测距和计算机图像处理相结合的木材表面缺陷检测方法。激光发射器发射扇形光源至传送台上的试件表面,从另一角度由相机对试件表面的激光光斑进行成像。通过图像处理,能自动识别裂纹及钝棱缺陷轮廓线,并得到外材面材宽和裂纹宽度尺寸信息。以7块含钝棱的毛边锯材和7块含裂纹的锯材进行试验,结果表明:在入射激光线与物镜光轴的夹角为60°的情况下,锯材外材面宽度和裂纹宽度检测值与实际值的误差均值都不超过±1 mm,证实了研究提出的锯材表面缺陷检测模型及构建的检测装置的高精度和可靠性。该技术能为木材加工中的自动优选下锯提供基础数据,可用于锯材优选自动化加工生产线。</br>In order to achieve the automatic online detection of wood surface defects, such as wane and crack, a new method based on laser triangulation and computer image processing for surface defects of sawn timber was proposed.The laser triangulation measurement was conducted by a laser beam focused on the object surface to be measured, and the laser spots on the timber surface were imaged from another angle.Through the image processing, the profile of wane and crack defects were identified and removed, the profile of external face was retained, and the information of the external face and crack width were obtained.The experiments were carried out with seven pieces of unedged sawn timber with wane and seven pieces of sawn timber with crack.The results showed that, when the incident angle of the laser beam was 60°, the average error of the measurement for external face and crack width were all not more than ±1 mm, indicating that the proposed detection model and the detection device for surface defects of sawn timber had high detection accuracy.This detecting model for surface defects of sawn timber can meet the requirements of the sawn timber detection, provide the basic data for the optimum sawing in the wood processing, improve the utilization ratio of wood resources, and can be directly used in the automatic and intelligent production line of wood processing
%K 激光三角测距
%K 木材表面缺陷
%K 在线检测
%K 图像处理
%K 霍夫变换&lt
%K /br&gt
%K laser triangulation
%K wood surface defects
%K on line detection
%K image processing
%K hough transformation
%U http://lkkf.njfu.edu.cn//oa/darticle.aspx?type=view&id=201706020