JPH09192986A - Automatic recognizing device for tool cutting edge position by image processing - Google Patents

Abstract

PROBLEM TO BE SOLVED: To precisely and automatically recognize the position of a cutting edge without removing a throw-away tip from a tool by providing a binary image data for discriminating the cutting edge part against the other part from the image data on the front of the tool. SOLUTION: The frequency distribution data corresponding to the tool center angle of a cutting edge picture element is formed by a frequency distribution data forming part 15. The arranging angle of the cutting edge around the tool is found by a cutting edge arranging angle calculating unit 19 from the frequency distribution data formed by the frequency distribution data forming part 15 by referring to the tool information stored in a tool information data base 17. The position of the tool cutting edge is specified by a tool cutting edge position calculating unit 21 from the cutting edge arranging angle found by the cutting edge arranging angle calculating part 19 by referring to the tool information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically recognizing a tool edge position of a cutting tool used in a machine tool,
In particular, the present invention relates to a method for automatically recognizing a tool edge position required when measuring the edge wear of a tool by image processing.

[0002]

2. Description of the Related Art As a method of measuring the wear of the cutting edge of a tool in a machine tool, there is known a method of measuring the cutting edge of the tool with an ITV camera in detail and measuring the wear of the cutting edge by image processing. "Automatic measurement of tool wear by image processing" (Journal of Precision Engineering, Vol.61, No.3)
(1995) pp. 368-371).

In this "automatic measurement of tool wear by image processing", in order to take a detailed picture of the tool cutting edge portion with the ITV camera, the work of positioning and arranging the tool at a position where the ITV camera can photograph the tool cutting edge portion is performed. , Manually performed by the operator.

For a tool such as a milling tool to which a throw-away tip is attached and used, conventionally, the throw-away tip is removed from the tool, and the throw-away tips are individually positioned at the photographing position by hand. Is done.

Further, in order to automate the above-mentioned positioning of the photographing position, "development of a non-contact tool damage monitoring device incorporating an autofocus mechanism by optical measurement" (Precision Engineering Society Vol. 55, No. 8 (1989)). 73-78), "Automatic measurement of drill wear by two ITV cameras and image processing" (Journal of Precision Engineering Vol. 56, N.
o. 10 (1990) pages 75 to 80), the throw-away tip is removed from the tool, the throw-away tip is arranged on the measurement table, and image data of the entire image is image-processed to extract the outer peripheral contour of the tip, A method of scanning the outer peripheral contour of each insert to find the minimum point of curvature and automatically recognizing the tip edge of the insert from the minimum point, or in a drill tool, extracting the outer peripheral contour from the front image of the drill tool A method is shown in which Hough transformation is performed on a point, two straight line components corresponding to a drill cutting edge portion are taken out, and the cutting edge position is automatically recognized.

[0006]

When the work of positioning and arranging the tool at a position where the tool cutting edge can be photographed by the ITV camera is manually performed by the operator, as long as this manual work is surely performed, the tool is surely performed. Although the cutting edge can be reliably photographed, it takes time and labor to set the imaging position of the tool cutting edge, which makes it impossible to fully automate the measurement of the tool wear. Also, regarding tools that can be used with the throw-away tip attached,
It takes time and effort to remove the throw-away chips and install the chips one by one, and if there are no abnormalities, it is necessary to remove, inspect, and attach the chips.

The method of removing the chips, arranging them on the measurement table, and recognizing the cutting edge portion from the curvature of the outer peripheral contour of the cutting chips obtained by image processing is easy to install the cutting chips, and the recognition of the cutting edge can be automated, and Multiple chips can be inspected, but for each chip lined up on the table, the outer peripheral contour is scanned to find the minimum point of curvature, and the cutting edge is determined from the number and distance of the calculated minimum points. Since the cutting edge recognition algorithm is complicated, it takes time to calculate. In addition, the measurement target is limited to the throw-away tip, it is poor in versatility, and even with this method, the tip has to be removed, so it takes time and effort, and for a tip with no abnormality, the same wasteful process as the above case must be performed. I have to.

Regarding the method of recognizing the tool cutting edge portion by the Hough transform, since the Hough transform extracts the linear component from the frequency distribution of the two-dimensional array of the distance ρ and the angle θ, it takes a long time to calculate and a large capacity. Need storage space.
Also, in the case of a drill tool, the linear component is limited to two cutting edge parts, but when it is necessary to extract a large number of linear components like a milling tool, the above-mentioned calculation time and storage capacity problems are involved. In addition, since the straight line components detected by the Hough transform do not necessarily indicate all cutting edge parts, an algorithm that determines which of the detected straight line components indicates a cutting edge part and which does not Will be required. For these reasons, this method has poor versatility and is not practical.

The present invention has been made by paying attention to the above-mentioned problems, and it is possible to remove the throw-away tip from the tool or the tool from the holder by the image processing of the cutting edge position of the cutting tool of the machine tool. The purpose is to provide a method of automatically and accurately recognizing the above.

[0010]

In order to achieve the above object, the invention according to claim 1 images a tool from the front by an image pickup means, and determines a cutting edge portion from image data of the tool front obtained by this image pickup. The binary image data that discriminates from other parts is obtained, and a polar coordinate system is set in which the pixel corresponding to the tool center is the origin in this binary image data, and the radial direction pixels are scanned at predetermined angles to cut edges. The pixel number of the density value indicating the part is counted for each predetermined angle, and the arrangement angle of the cutting edge part around the tool center is obtained from the frequency distribution data corresponding to the tool center angle of this cutting edge pixel, and this angle data and the tool diameter The tool edge position is specified from the data.

According to a second aspect of the present invention, in the method for automatically recognizing a tool edge position according to the first aspect, data relating to a rotational direction of a tool is stored in a tool information database in advance, and based on the tool rotational direction data, In the frequency distribution data, either the angle at which the frequency of the cutting edge portion pixels sharply increases or the angle at which the frequency number sharply decreases is selected, and the arrangement angle of the cutting edge portions is obtained.

The invention according to claim 3 is the invention according to claim 1 or 2.
In the method for automatically recognizing the tool edge position described in (1), the tool diameter data is stored in advance in the tool information database, and the tool diameter data is acquired from the tool information database.

According to a fourth aspect of the invention, in the method for automatically recognizing a tool edge position according to any one of the first to third aspects, the image pickup means of the image pickup means is arranged such that the central optical axis of the image pickup means and the rotation center of the tool coincide with each other. The placement position and the imaging position of the tool are defined,
The pixel corresponding to the tool center is specified based on this positional relationship.

In the method for automatically recognizing the tool blade edge position according to these inventions, the blade edge portion of the tool is obtained from the image data of the entire front surface of the cutting tool of the machine tool (for the tool to which the throw-away tip is attached, the tip is still attached). Image data for discriminating the cutting edge portion from other portions, that is, binary image data obtained by extracting only the cutting edge portion is obtained by utilizing the difference in gray value between the Set the r-θ polar coordinate system with the pixel at the center of the tool as the origin, scan the pixels in the radial direction at every predetermined angle, and calculate the frequency distribution in which the density value indicating the cutting edge portion, for example, the pixel with the density value 1 is counted. Ask. The edge position of the tool is determined based on the frequency distribution thus obtained and the tool information such as the tool type, the tool diameter, and the number of cutting edges.

[0015]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 shows a tool image pickup device used for carrying out the method for automatically recognizing the tool edge position according to the present invention. In FIG. 1, 1 is a tool holder installation jig that positions and holds a tool (drill tool) T together with a tool holder 3 at an imaging position, and 5 is an image of the entire tool T attached to the tool holder installation jig 1 from the front. ITV for whole tool imaging
The camera, 7 is the tool T from the automatically recognized tool edge position.
Each of the close-up ITV cameras that captures a detailed image of the blade edge is shown.

The arrangement position of the ITV camera 5 and the arrangement (imaging) position of the tool by the tool holder installation jig 1 are determined so that the central optical axis of the ITV camera 5 and the rotation center of the tool T coincide with each other. . In this case, if the tool holder installation jig 1 is fixedly arranged, the ITV camera 5 determines the coordinate position in the X axis direction and the Y axis direction in FIG. 1 in the X axis direction and the Y axis direction of the tool holder installation jig 1. It is only necessary that the coordinate position is fixed and the ITV camera 5 can move only in the Z-axis direction as needed.

FIG. 2 shows an embodiment of an automatic tool edge position recognizing apparatus used for implementing the automatic tool edge position recognizing method according to the present invention.

This tool edge position automatic recognition device is an ITV
An A / D converter 11 for converting an analog image signal output from the camera 5 into a digital image signal (image data);
The image data from the D / D converter 11 is discriminated between the cutting edge portion (density value 1) and the other portion (density value 0) by utilizing the difference in gray value between the cutting edge of the tool and the other portion. Binarization processing unit 13 for converting into binary image data to be processed, and binarization processing unit 13
In the binary image data obtained by setting a polar coordinate system with the pixel corresponding to the tool center as the origin, the radial direction pixels are scanned at every predetermined angle, and the number of pixels of the density value indicating the cutting edge is determined at every predetermined angle. The frequency distribution data generation unit 15 that generates frequency distribution data corresponding to the tool center angle of the cutting edge pixel, the tool type of the tool to be inspected, the tool diameter,
Tool information database 17 that previously stores tool information such as the number of cutting edges and rotation direction, and tool information database 1
The cutting edge portion arrangement angle calculation unit for obtaining the arrangement angle of the cutting edge portion around the tool center from the frequency distribution data generated by the frequency distribution data generation unit 15 with reference to the tool information (rotational direction data) stored in 7. 19 and the tool information (tool diameter data) stored in the tool information database 17 to identify the tool edge position from the cutting edge placement angle obtained by the cutting edge placement angle calculation unit 19 It has a calculation unit 21.

Next, a procedure for carrying out the automatic tool edge position recognizing method according to the present invention by the automatic tool edge position recognizing apparatus having the above-mentioned structure will be described.

(A) The tool T is attached to the tool holder installation jig 1 together with the tool holder 3 while the tool T is still attached to the tool holder 3, and the ITV camera 5 facing the tool T images the entire tool T from the front.

In the case of a tool with a throw-away tip attached, the entire tool T is imaged from the front by the ITV camera 5 with the tip still attached.
FIG. 3 shows a monitor display example of an image based on multi-valued image data obtained by imaging the milling tool from the front.

(B) Since the cutting edge portion b of the tool T appears to be relatively brighter than other portions of the tool, the threshold value in the binarization processing unit 13 is utilized by utilizing the difference in density value. A value is set to make the image data binary image data. In this case, the pixel of the tool cutting edge portion b has a density value of 1 and is displayed in white on the monitor.

When the binary image data has a lot of noise components (dust components), labeling processing is performed on the binary image data and the tool characteristics such as the tool type and the number of cutting edges obtained from the tool information, Only the tool cutting edge can be extracted from the area and center position of each label. FIG. 4 shows multi-valued image data for displaying the image shown in FIG.
An example of a monitor display in which only the cutting edge portion b is extracted by binarizing and removing the noise component is shown.

Since the tool T and the center axis position of the ITV camera 5 coincide with each other, the center position of the tool T on the image data obtained by photographing with the ITV camera 5 is always a constant pixel (x0, y0). ).

(C) For binary image data obtained by extracting the cutting edge portion b as illustrated in FIG. 4, the pixel (x0, y0) is set as the origin as shown in FIG. r-θ
Set the polar coordinate system of.

The x-coordinate value and the y-coordinate value when the radius is r and the angle is θ are obtained by the equations (1) and (2).

X = x0 + rcosθ (1) y = y0 + rsinθ (2) Then, the pixel is scanned in the radial direction at every predetermined angle to obtain a pixel having a density value indicating the cutting edge portion b, in this case, The pixels having the density value of 1 are counted, and the frequency distribution data of the cutting edge part pixels as shown in FIG.
Create by. This frequency distribution data is frequency distribution data corresponding to the tool center angle. Since the angle θ is taken as an angle in the counterclockwise direction as shown in FIG. 4, the horizontal axis (angle The right direction (deployment direction) of θ) and the counterclockwise direction match.

When the binary image data of the image illustrated in FIG. 4 is scanned on the r-θ polar coordinate system, the cutting edges are distributed on the circumference, so that the frequency is high in the cutting edge portion b. Therefore, in the frequency distribution data shown in FIG. 6, the cutting edge portion b appears as a peak portion.

(D) Since the rotating direction of the tool is known from the tool information stored in the tool information database 17, for example, in the case of a tool rotating counterclockwise, the required arrangement angle of the cutting edge portion b is This is a point (angle) at which the frequency of the peak portion in the frequency distribution data sharply decreases from a high value to a low value in consideration of how to take the frequency distribution data in the horizontal axis direction.

At this point, the frequency distribution data is differentiated to obtain a point where the differential value becomes higher on the − (minus) side, or the threshold value is determined from the maximum value and the minimum value of the frequency number, and the threshold value is set. +
It can be determined by a method such as obtaining a point crossing from the (plus) side to the- (minus) side.

As a result, the cutting edge placement angle calculation unit 19 refers to the tool rotation direction data stored in the tool information database 17 to obtain the above-mentioned point, that is, the placement angle of the cutting edge b.

(E) The tool diameter is the tool information database 17
Since it is obtained from the tool information stored in, the tool edge calculating unit 21 calculates from the tool diameter acquired from the tool information database 17 and the arrangement angle of the cutting edge b obtained in (d),
The position of the cutting edge can be determined in the polar coordinate system, and the x-coordinate value and the y-coordinate value of the cutting edge can be obtained by converting the coordinate into the rectangular coordinate system.

(F) As a method for obtaining the Z-coordinate value of the tool cutting edge, and if necessary, a method for obtaining the attachment angle of the insert such as a milling tool or the tip angle of the drill tool, the following method is used. is there. The tool is photographed from two locations with an ITV camera, and it is determined from the principle of triangulation. It is obtained optically from the focus position of the ITV camera when the front of the tool is photographed. A secondary sensor such as a laser displacement meter is used. The tool length, tip mounting angle, and drill tip angle are registered in advance in the tool information database and determined from the tool information. The most realistic method of ~ is the method of.

By the above method, the X, Y, Z coordinates of the tool cutting edge are obtained, and thereafter the close-up camera 7 for photographing the cutting edge of the tool is moved to the obtained cutting edge position, and the wear amount is measured by normal image processing. Can measure the amount of tool wear.

FIGS. 7, 8 and 9 are examples of monitor display by tool front image data, example of monitor display by binary image data, and frequency distribution graph when the same steps as those described above are performed on a drill tool. Are shown respectively. Also in this case, the angle of the cutting edge can be determined by finding the point where the frequency number sharply decreases from the frequency distribution data.

The tool information such as the tool type, the tool diameter, the number of cutting edges, and the rotation direction may be input from the outside in addition to being read from the database.

In the above, the present invention has been described in detail with respect to a specific embodiment, but the present invention is not limited to this, and various embodiments are possible within the scope of the present invention. It will be apparent to those skilled in the art that it is possible.

[0039]

As can be understood from the above description, according to the tool edge recognizing method of the present invention, it is not necessary to remove the tool or the throw-away tip from the tool holder or the tool, and the axes of the tool and the ITV camera coincide with each other. As described above, the tool edge position can be automatically recognized simply by setting the positional relationship between the two, so that the work is simple and the cycle time can be shortened.

Further, since the position of the cutting edge of the tool can be recognized without removing the tool or the throw-away tip from the tool holder or the tool, by combining the close-up camera for photographing the cutting edge with a device that can move to the desired position, for example, a machine tool The above-mentioned device can be installed in the tool magazine section of (1) to automatically measure tool wear.

Moreover, the angle of the tool cutting edge is determined by a simple method in which the photographed binary image data of the front surface of the tool is used to scan the polar coordinate system with the tool center pixel as the origin to obtain the frequency distribution. The same algorithm can be used to automatically recognize the tool tip position with tools that use away tips and tools that do not, and the program is simple, the calculation does not take time, and the storage capacity is small.

Further, since the blade edge position of the tool is obtained from the blade edge angle obtained by the image processing and the characteristic peculiar to the tool obtained from the tool information, the precision is high and the reliability is high.

[Brief description of the drawings]

FIG. 1 is a schematic view of a tool photographing device used for carrying out a tool edge recognition method according to the present invention.

FIG. 2 is a block diagram showing one embodiment of an automatic tool edge position recognizing device used for implementing the tool edge position recognizing method according to the present invention.

FIG. 3 is a photograph showing an example of a monitor display by multi-valued image data on the front of the milling tool.

FIG. 4 is a photograph showing an example of a monitor display by binary image data on the front of the milling tool.

FIG. 5 is an explanatory diagram showing the concept of a method of obtaining a frequency distribution by scanning in a polar coordinate system.

FIG. 6 is a frequency distribution graph of cutting edges in a milling tool.

FIG. 7 is a photograph showing a monitor display example based on multi-valued image data of the front surface of a drill tool.

FIG. 8 is a photograph showing a monitor display example based on binary image data of the front surface of a drill tool.

FIG. 9 is a frequency distribution graph of cutting edges in a drill tool.

[Explanation of symbols]

 1 Tool holder installation jig 3 Tool holder 5 ITV camera for whole tool imaging 7 Close-up ITV camera 11 A / D converter 13 Binarization processing unit 15 Frequency distribution data generation unit 17 Tool information database 19 Cutting edge part placement angle calculation 21 Tool edge position calculator

Claims (4)

[Claims] 1. A tool is imaged from the front by an imaging means, binary image data for discriminating a cutting edge portion from other parts is obtained from image data of the tool front obtained by this imaging, and in this binary image data Set a polar coordinate system with the pixel corresponding to the tool center as the origin, scan the pixels in the radial direction at every predetermined angle, and count the number of density value pixels indicating the cutting edge at each predetermined angle. The placement angle of the cutting edge around the tool center is obtained from the frequency distribution data corresponding to the tool center angle of pixels, and the tool edge position is specified from this angle data and the tool diameter data. Automatic recognition method. 2. Data relating to the rotational direction of a tool is stored in advance in a tool information database, and based on this tool rotational direction data, the angle at which the frequency of cutting edge pixel pixels sharply increases or abruptly increases in the frequency distribution data. The tool edge position automatic recognition method according to claim 1, wherein any one of the decreasing angles is selected and the arrangement angle of the cutting edge portion is obtained. 3. The tool diameter data is stored in advance in the tool information database, and the tool diameter data is acquired from the tool information database.
A method for automatically recognizing the tool edge position described in. 4. The arrangement position of the image pickup means and the image pickup position of the tool are set so that the central optical axis of the image pickup means and the rotation center of the tool coincide with each other, and the pixel corresponding to the tool center is specified from this positional relationship. The automatic recognition method of the tool edge position according to any one of claims 1 to 3, characterized in that.