基于计算机视觉的刀具在机检测

发布时间：2018-04-22 15:59

本文选题：刀具检测 + 自动对焦　；参考：《天津大学》2016年硕士论文

【摘要】：随着科学技术和社会生产的不断发展,对工件的质量和生产率提出了越来越高的要求。在加工过程中,刀具的状态直接决定了工件的质量,对刀具磨损进行监测,及时把刀具的磨损信息反馈给机床的控制系统,可以避免造成工件报废和机床的损害;通过对刀具几何形状的检测,可以调整机床加工参数,补偿加工轨迹,从而实现精密加工。针对刀具在机监测和检测需求,搭建了图像采集系统,并利用该平台进行了自动对焦算法、刀具磨损监测和刀具几何形状检测等方面的研究。1.根据刀具在机检测系统的需求,选择合适类型和参数的相机、镜头、光源等设计并搭建图像采集系统。在此基础上,为了确保刀具位于相机的焦平面处,通过对比几种对焦评价函数的对焦灵敏性并进行性能优化,最后采用Brenner梯度函数结合超精密机床各轴的移动来实现自动对焦,通过先粗调后微调的调焦方式,加快了对焦过程,实现了高质量图像的采集。2.对基于灰度值的模板匹配方法和基于形状的模板匹配方法进行了研究,从匹配原理、系统稳定性和运算效率三个方面对两种匹配方法的适用性进行分析,并将金字塔算法应用于模板匹配过程中,实现了刀具磨损的快速、稳定监测。3.自主设计圆弧刃刀具几何形状检测算法,并从算法效率、稳定性和精度方面与轮廓分割圆弧拟合算法进行对比。结果表明,自主设计算法检测稳定性更佳,并且刀具形状较差或者刀具磨损出现豁口时都能实现精确检测。4.针对超出视野范围的圆弧刃刀具,为了保证检测精度,提出了基于刀具轮廓的拼接算法。采用轮廓识别、仿射变换、边缘融合等技术保证了刀具轮廓的准确拼接,最后验证了该拼接算法的检测精度。
[Abstract]:With the development of science and technology and social production, the quality and productivity of workpiece are demanded more and more. In the process of machining, the state of the tool directly determines the quality of the workpiece. Monitoring the tool wear and feeding back the tool wear information to the control system of the machine tool in time can avoid the scrap of the workpiece and the damage of the machine tool. By measuring the geometric shape of the tool, the machining parameters can be adjusted and the machining track can be compensated, thus the precision machining can be realized. In order to meet the needs of tool monitoring and inspection, an image acquisition system is built, and the automatic focusing algorithm, tool wear monitoring and tool geometric shape detection are studied by using the platform. According to the requirements of the tool in machine detection system, the image acquisition system is designed and built by selecting the appropriate type and parameters of camera, lens, light source and so on. On this basis, in order to ensure that the tool is located at the focal plane of the camera, the focus sensitivity of several focusing evaluation functions is compared and the performance is optimized. Finally, Brenner gradient function combined with the movement of each axis of the ultra-precision machine tool is used to realize the automatic focusing. The focusing process is accelerated and the high quality image acquisition is realized by the focusing mode of coarse tuning and fine tuning. The template matching method based on gray value and the template matching method based on shape are studied. The applicability of the two matching methods is analyzed from three aspects: matching principle, system stability and operation efficiency. The pyramid algorithm is applied to the template matching process, and the tool wear is monitored quickly and stably. 3. The geometric shape detection algorithm of arc cutting tool is designed and compared with the contour segmentation arc fitting algorithm in terms of efficiency, stability and precision. The results show that the self-designed algorithm is more stable and can detect accurately when the tool shape is poor or the tool wear is gaping. In order to ensure the detection accuracy, an algorithm based on cutter contour is proposed for the arc edge cutter which is out of sight. The contour recognition, affine transformation and edge fusion techniques are used to ensure the accurate stitching of the tool contour. Finally, the detection accuracy of the proposed algorithm is verified.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TG71;TP391.41

【相似文献】