RESEARCH ON THE DETECTION OF CUTTING TOOL WEAR BY MACHINE VISION
Mei Anhua Wang Jinhui
Optical Instrument Department, W. T. V of Surving and Mapping
This paper presents a method of applying a system which consists of an Optoelectronic scanner(TV camera or CCD) and
an IBM/pc computer to obtain the micrograph information from the worn region on the flank of the cutting part of the tool
and analyze and process the picture by means of imagine processing and mode recognizing technology. The on-line and off-
line methods are used respectively to calculate accurately the worn value, recognize the form of wearing and forecast the
tool life.At the last, the optimum valued of cutting velocity and feed are given, which meet the principle of the lowest cost
and longest tool life It is proved by experiments that the conclusion of the theoretical analysis and calculation agrees
with the practical test, Thearticle provides a practical and precise way of checking and controling cutting tools on Line.
Keyword: Machion Vision, Cutting tool, Wear
1.0btaining the wear imagine on the flank of 2.2 Extracting the edge of wear region
cutting tool ;
Cutting tool will be worn after being used. The According to the character of geometric state and
worn region appears around the corner. There are two the value of the wear, the wear type could be
kinds of wear: the face wear and the flank wear. In recognized, The precision and reliability of
theory and practice, the flank wear is widely used. recognition relys on the measuring precision for the
So, in this research, we collect, process and wear geomety. This research improves mesuring
analyce the flank wear imagine of cutting tool. The precision by means of making binary image, extracing
system which has been used in the reseach shows in the edge of worn region on binary picture and
Finl.1. gauging the edge.
Selecting threshold is a impartant step of making
binary image. When the wear picture is being taken
by microscope and TV camera, the flank’s wear area
L ht Source is on the centre of the field of microscope, In the
i picture, there are three parts: worn area, unworn
area of the flank and other's parts(back ground) .
T V There are deferent gray level in three prats.
x | WwüCx, y? is defined to the gray level function of
Camera. Microscope background. ww0(x,y) is distributed equally over the
background area Cw0). ;
The unworn part of the flank reflects partly light
E ud and shines higher than the background when lighted
m en by incandestent lamp. We definite it’s gray level
P Picture Cutting fnuction as Gwl(x,y), which is distributed over the
286 Monitor area, Showed as wi,
Tool The worn prat of flank reflect light strongest in
the three parts of a picture, It’s gray level
function is Hw2(x,y), which is distribute over the
X area showed by w2.
3 ^f Level function of the image can be described as:
W (x,y) x, yEw0
t Gvpz(6 esp xjy€wl (2-D
H GOGy) xy€w2
The system consists of TV Camera, microscope(C30^, Obviously,
Computer and picture monitor. The micrograph has HCx, y)>GCx, y)>WCx, y) (2-2)
been obtained by the system, which is showed Fig, lof Therefore, the gray level histogram of the image
the appendix, In the micrograpl, the worn region is is showed as Fig. 1.2
closed and irregular. Because of using incandescent
lamp for lighting, light affects directly the
clarity of the picture and gauging precise, In order
to prove gauging precise, the light source must be
adjust carefully to make high reflecting Light on
wear region so the distingush in worn region and
unworn region are very clear, the imagine must be
reprocessed with smoothing and filtering methods,
before being gauged.
———
2.Preprocessing micrograph and extracting the edge,
Number of Points
2.1 Preprocessing:
Because incandescent is not distriluted equally
over the flank of cutting tool, the quality of the
pictures are affected and much disturbance takes Gray Value
place in the picture, So the picture must be
processed by mediom filter method to cancel the
disturbance and make the edge of the wear region From these, we can choice the threshold for making
clear, After filtering, the gray level histogram can binary image, The binary image has been made that
be made to prepare choising the thresholding, shows as the Fig.2 in appendix,
which is the key of making binary picture, The edge of the worn area can be produced by the
methods of scanning, extracting and thinning. The