基于机器视觉的房屋倾斜测量方法  

A METHOD FOR MEASURING THE INCLINATION OF HOUSES BASED ON MACHINE VISION

作　　者：赵广辉[1] 王剑[2] 范宇鑫 赵启林 袁辉 Zhao Guang-hui;Wang Jian;Fan Yu-xin;Zhao Qi-lin;Yuan Hui

机构地区：[1]昆山市建设工程质量检测中心,江苏昆山215337 [2]南京工业大学,南京211516 [3]浦口区科技成果转化服务中心,南京210000

出　　处：《建筑技术开发》2022年第21期136-139,共4页Building Technology Development

摘　　要：为了精确测量房屋的倾斜程度,提出基于机器视觉的倾斜测量方法。利用固定在建筑物上的激光器向靶标发射激光,通过计算激光光斑的位移,来确定建筑的倾斜位移。对此,使用透视投影变换对获得的光斑图像进行校正处理,以应对房屋倾斜后带来的误差,然后用最大类间方差法对图像进行分割,通过灰度重心法提取十字标记与激光斑点的中心,以获得激光光斑中心的位置,并实时监测得到光斑的位移,三角函数换算得到房屋的真实倾斜位移。实验结果表明,所提方法能将激光光斑中心定位的误差稳定在5%以内,能够满足现场测量要求。In order to accurately measure the inclination of houses,an inclination measurement method based on machine vision is proposed.The laser fixed on the building is used to emit laser light to the target,and the tilt displacement of the building is determined by calculating the displacement of the laser spot.In this regard,this paper uses perspective projection transformation to correct the obtained spot image to deal with the error caused by the tilt of the house,and then uses the maximum inter-class variance method to segment the image,and extracts the cross mark and laser spot by the gray centroid method.The center of the laser spot can be obtained to obtain the position of the center of the laser spot,and the displacement of the spot can be obtained by real-time monitoring.The trigonometric function is converted to obtain the real tilt displacement of the house.The experimental results show that the proposed method can stabilize the error of laser spot center positioning within 5%,which can meet the requirements of on-site measurement.

关 键 词：机器视觉 激光光斑 最大类间方差法 灰度重心法 

分 类 号：TU198.2[建筑科学—建筑理论]