A Study on Automatic On-machine 3D Modeling System of Cutter and Holder

Researcher: Xi ZhangXiZhang.jpg

Start date:    January 2007

End date:      December 2009

Keyword:      3D Model,Cutting tools,

                     Singe View Reconstruction


To study the feasible methods for accurate 3D model reconstruction of cutting tools and develop the prototype system.


The surface of revolution (SOR) model of a cutting tool as well as tool offset information is required by machining simulation and setup. In order to guarantee the reliability and efficiency of machining, it is necessary to obtain both information on-machine at the same time once the cutter is clamped on the spindle. However, currently there is no such measuring methods and devices. In order to solve this problem, this research presents a new tool measuring and setting system, which can obtain both SOR model and tool offset on-machine rapidly. Only one camera is installed on the machine tool. An automatic on-machine camera calibration procedure is designed to quickly calculate the position and orientation as well as the optical property the camera. Then the image of a rotating cutter is taken by the camera. Since the cutter is not a real solid SOR due to flutes and cutting edge, a complete image is composed by overlapping the image sequence of a rotating cutter with backlighting. From this composed image, the contour of the cutter is segmented and classified. One image point on the contour is used to calculate one-cross section of the SOR model. Therefore the SOR model can be represented as a stack of cross sections, from which the generating function of SOR can be extracted. Finally tool radius offset and length offset can be obtained from the SOR model. The prototype system has been designed and implemented on machine tool. The SOR model can be built within 5 seconds. The measurement accuracy of vision system has been verified by Blum laser tool setting system as around 35 microns if a 8mm lens and 1 mm extensions are used.


  1. 3D Reconstruction and pose determination of the cutting tool from a single view.
    Zhang, X., et al.
    2007. Lake Tahoe, NV, United states: Springer Verlag.
  2. Automatic segmentation of the apparent contour for 3D modeling of cutting tools from single view.
    Zhang, X., et al.
    2008. Las Vegas, NV, United states: Springer Verlag.


If you have a pdf reader, please look at the poster.
The poster is pdf file. The pdf file will be opened by clicking.