KR20180046267A - Distinguishing method of tool class using vision and distinguishing system thereof - Google Patents

Abstract

The present invention relates to a method and a system for distinguishing a tool type. More specifically, the present invention relates to a method and a system for distinguishing a tool type using a vision, which capture a tool mounted on a tool post to distinguish a type of tool from a tool image, transmit information of the distinguished tool to a machine tool and allow the machine tool to automatically perform a process. The system for distinguishing a tool type comprises: a camera capturing the image of the tool installed in the tool post; and a computer in which a computer program is installed, and which receives the tool image from the camera to distinguish the tool type.

Description

TECHNICAL FIELD The present invention relates to a tool discrimination method and a tool discrimination method using vision,

The present invention relates to a tool type discriminating method and a tool type discriminating system, wherein a tool mounted on a tool stand is photographed to discriminate the type of tool from the tool image, and the information of the discriminated tool is transmitted to a machine tool, The present invention relates to a tool type discrimination method and a tool type discrimination system using a vision,

In order to perform machining with a machine tool, various parameters of the tool to be machined must be input directly to the machine tool by the operator.

When the information of the tool is input directly to the machine tool by the operator, it is possible to generate a wrong input, and since the type of the tool must be confirmed and input, it takes a long time to input the tool information, have.

Korean Patent No. 10-1013749 discloses a technology of a CNC machine tool equipped with a vision system.

This CNC machine tool has the advantage of improving precision and simplicity of control by controlling the machine tool by acquiring the machining process for the workpiece as an image.

However, the conventional CNC machine tool does not discriminate the type of tool but uses a vision system to confirm the processing state of the workpiece, so that there is a problem that the type of tool can not be discriminated.

In addition, since the conventional CNC machine tool uses a template matching method for image processing, there is a problem that the processing state can not be detected for images other than the learned data.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to provide a tool type discrimination method capable of automatically inputting tool information into a machine tool by discriminating a tool type from an image of a tool mounted on the tool table And a tool type discrimination system.

Another object of the present invention is to provide a tool type discrimination method capable of accurately detecting the tool flute length, neck length, shank length, neck diameter and bag diameter, And a tool type discrimination system.

According to an aspect of the present invention, there is provided a method of manufacturing a tool, comprising: capturing an image of a tool mounted on a tool holder using a camera (hereinafter, referred to as a tool image); The computer receiving the tool image and extracting an outline image of the tool by detecting an edge of the tool; Calculating a flute length, a neck length, a shank length, a neck diameter or a bag diameter of the tool from the contour image; And discriminating the type of the tool by inputting the blade length, the neck length, the bag length, the neck diameter, or the bag diameter into a data classifier.

In a preferred embodiment, the step of extracting the outline image includes: receiving the tool image and detecting the outline of the tool; Confirming whether the contour is a closed curve; Extracting an image of the contour line as the contour line image if the contour line is a closed curve; And when the contour is not a closed curve, changing the shooting conditions to retake the contour image, detecting the contour of the tool from the re-captured tool image and merging it with the contour of the previously detected tool, And returning to the step of confirming whether or not it is possible.

In a preferred embodiment, the tool image is an image photographed such that the blade of the tool faces downward, or the tool is rotated in a state in which the blade is oriented in the lateral direction, and then the blade is rotated downward. Calculating the neck length, the bag length, the neck diameter, or the bag diameter comprises: detecting a bottom pixel in the contour image; Calculating a width of the outline image from the height of the bottom pixel (hereinafter, referred to as 'first height') to the top; Detecting a height (hereinafter, referred to as " second height ") at which the width change rate of the outline image is less than or equal to a first threshold change rate; Detecting a height (hereinafter, referred to as a 'third height') at which the width change rate of the outline image becomes equal to or greater than a second threshold change rate after the second height detection; And detecting a top pixel of the contour image, wherein the blade length is calculated as the distance between the first height and the second height, and the neck length is between the second height and the third height The bag length is calculated as a distance between the third height and the height of the uppermost pixel (hereinafter, referred to as 'fourth height'), and the neck diameter is calculated as the distance of the contour image on the second height And the bag diameter is calculated as the width of the contour image on the third height.

In a preferred embodiment, the data classifier is a Neural Network Classifier.

In a preferred embodiment, when the type of the tool is determined in the neural network classifier, data on the determined tool length, the neck length, the bag length, the neck diameter, or the bag diameter is transmitted to the tool, And returning to the step of photographing the tool image.

The present invention further provides a computer program stored in a medium for functioning as a computer to perform the tool type discrimination method.

The present invention also relates to a storage medium on which the computer program can be stored; And a communication medium capable of transmitting a computer program stored in the storage medium through a communication network.

The present invention also relates to a camera for photographing an image of a tool installed on a tool stand; And a computer installed with the computer program and receiving a tool image from the camera and determining a tool type.

The present invention also relates to a bed comprising: a bed; A tool base provided on the bed and capable of machining a workpiece by mounting a tool; And the tool type discrimination system for discriminating a tool type by photographing a tool of the tool base.

The present invention has the following excellent effects.

According to the tool type discriminating method and the tool type discriminating system of the present invention, it is possible to detect the tool flute length, neck length, shank length, neck diameter and bag diameter from the image of the tool, It is possible to input the information of the tool to the machine tool automatically and accurately.

According to the tool type discriminating method and the tool type discriminating system of the present invention, since the neural network classifier is used for classifying the tool type, it is possible to discriminate the tool type with a small amount of data. As the data increases, There is an advantage to be improved.

1 is a view showing a tool type discrimination system according to an embodiment of the present invention;
FIG. 2 is a flowchart of a tool type determination method according to an embodiment of the present invention;
FIG. 3 is a flowchart for explaining an outline image extraction process in a tool type determination method according to an embodiment of the present invention;
4 is a diagram for explaining a method of extracting tool information in a tool type determination method according to an embodiment of the present invention.

Although the terms used in the present invention have been selected as general terms that are widely used at present, there are some terms selected arbitrarily by the applicant in a specific case. In this case, the meaning described or used in the detailed description part of the invention The meaning must be grasped.

Hereinafter, the technical structure of the present invention will be described in detail with reference to preferred embodiments shown in the accompanying drawings.

However, the present invention is not limited to the embodiments described herein but may be embodied in other forms. Like reference numerals designate like elements throughout the specification.

FIG. 1 illustrates a tool type determination system for performing a tool type determination method according to an embodiment of the present invention. The tool type determination system 100 includes a camera 110 for capturing an image of the tool 10, And a computer (120) for determining the type of tool from the tool image photographed by the camera (110).

Further, the computer 120 inputs the information of the discriminated tool to the machine tool, and the tool rest 20 of the machine tool drives the tool 10 to machine the work.

Further, the machine tool includes a bed (not shown) as a support and a tool table 20 provided on the bed, and the present invention can be applied to a finished product including the tool type discrimination system 100 May be provided in the form of a machine tool.

In addition, the computer 120 is provided with a computer program for performing the tool type determination method of the present invention, and the computer program functions as the computer 120 to perform the tool type determination method of the present invention.

In addition, the computer 120 may be various computing devices capable of image processing in addition to a general personal computer, and may be provided in the form of an embedded system.

In addition, the computer program may be stored in a separate recording medium, and the recording medium may be designed and configured specifically for the present invention or may be known and used by those having ordinary skill in the computer software field .

For example, the recording medium may be a magnetic medium such as a hard disk, a floppy disk and a magnetic tape, an optical recording medium such as a CD and a DVD, a magneto-optical recording medium capable of serving also as magnetic and optical recording, Or the like, or a hardware device specially configured to store and execute program instructions by itself or in combination.

In addition, the computer program may be a program consisting of program commands, local data files, local data structures, etc., alone or in combination, and may be executed by a computer using an interpreter or the like as well as machine code Lt; RTI ID = 0.0 > language code. ≪ / RTI >

In addition, the computer program may be stored in a server system capable of transmitting data through a communication network, and the computer may download and install the computer program from the server system.

The server system is a computer system comprising a storage medium storing the computer program, and a communication medium capable of transmitting the computer program stored in the storage medium through a communication network.

Hereinafter, a tool type determination method according to an embodiment of the present invention will be described in detail with reference to FIG. 2 to FIG.

Referring to FIG. 2, first, the camera 110 captures an image of the tool 10 mounted on the tool rest 20 (S1000).

In addition, the camera 110 photographs an image so that the blade of the tool 10 faces downward, and inputs the captured tool image to the computer 120.

Next, the computer 120 converts the tool image into an outline image and extracts it.

In this case, the computer 120 rotates the tool image and rotates the tool 10 so as to rotate the tool 10 in a direction perpendicular to the tool 10, It is possible to extract an outline image after facing downward.

More specifically, referring to FIG. 3, the computer 120 receives the tool image, preprocesses the tool image, and detects the contour of the tool.

Here, the pre-processing is a preprocessing process for contour detection, for example, converting the tool image into a monochrome image, and removing noise to normalize the image.

In addition, contour detection of a tool can utilize a known contour detection technique such as a Soble mask, a Canny edge detection technique, and the like.

Next, it is confirmed whether the detected contour is a closed curve (S2200).

If it is a closed curve, the detected contour is extracted as an outline image.

However, when the image is not a closed curve, since the image is not properly captured due to noise, the brightness of the illumination is adjusted or the position of the tool 10 is adjusted, and then the tool image is taken again (S2400).

Next, an outline of the re-photographed tool image is detected (S2500), and the outline is merged with the previously detected outline (S2600).

Next, the process returns to step S2200 of determining whether the contour is a closed curve, and whether the contour is a closed curve. If the closed curve is a closed curve, the corrected contour is extracted as the contour image (S2300).

However, if the corrected contour is not a closed curve, the process of re-photographing the tool image and merging the contour until the contour becomes a closed curve is repeated.

Referring back to FIG. 1, when the extraction of the outline image is completed (S2000). The blade length, neck length, bag length, neck diameter and bag diameter of the tool are calculated from the contour image (S3000).

4 shows an extracted contour image 200 wherein the tool 10 includes a blade 210 that is in direct contact with the workpiece, a neck 220 that supports the blade 210, And a bag 230 extending to the neck 220 and fixed to the tool rest.

In addition, the blade 210 has a peak at its tip, and the neck 220 and the bag 230 are cylindrical and appear as a rectangle in the contour image 200.

The diameter of the neck 220 is larger than the diameter of the bag 230.

Therefore, the blade length, the neck length, the bag length, the neck diameter and the bag diameter of the tool are calculated in consideration of the characteristics of these tools.

First, the lowermost pixel (a) is detected in the contour image.

The lowest pixel (a) is the end of the blade 210.

Next, the horizontal width (w) of the outline image is calculated while climbing upward from the height (h, first height) of the lowermost pixel (a) in the outline image.

Next, the width change rate is calculated from the difference between the currently calculated width and the previously calculated width, and the height (h2, second height) of the outline image in which the calculated width change rate is less than or equal to the first threshold change rate is detected.

In addition, since the blade 210 has a generally triangular shape, there is a rate of change in width. In the case of the neck 220, the blade 210 is rectangular.

Accordingly, the point (b) at which the width change rate becomes '0' or becomes slower than the first threshold change rate can be detected as the lowermost part of the neck 220.

Next, the width change rate is calculated while moving upward from the second height h2, and when the width change rate is greater than or equal to the second threshold change rate, the height is detected as the third height h3.

In addition, the third height h3 is a portion (c) where the bag 230 is connected to the neck 220, and thus the lateral width of the contour image is sharply reduced.

That is, the third height h3 can be detected using this feature.

Next, the uppermost pixel d of the contour image is detected, and the height of the uppermost pixel d is detected as the fourth height h4.

Next, a blade length, a neck length, a bag length, a neck diameter, and a bag diameter are calculated based on the first through fourth heights h1, h2, h3, and h4 and the width of the outline image.

The blade length is a distance between the first height h1 and the second height h2 and is calculated by subtracting the first height h1 from the second height h2.

Next, the neck length is calculated by subtracting the second height h2 from the third height h3, and the bag length is calculated by subtracting the third height h3 from the fourth height h4.

Next, the neck diameter w1 is calculated as the width of the outline image at the second height h2, and the bag diameter w2 is calculated as the width of the outline image at the third height h3.

Therefore, the blade length, the neck length, the bag length, the neck diameter, and the bag diameter of the tool 10 can be accurately calculated from the contour image 200 and derived from the tool information.

Referring again to FIG. 2, the type of tool is determined by inputting the calculated blade length, neck length, bag length, neck diameter, and bag diameter into a data classifier.

In addition, in the present invention, the data classifier is a neural network classifier, and the tool type can be discriminated even with a small amount of data, and the accuracy of discrimination is greatly increased when data is increased.

Next, when the tool type is identified, the information of the identified tool is transmitted to the tool rest 20 to perform the machining. Otherwise, the process returns to the step S1000 of photographing the tool image (step S1000) .

The tool data transmitted to the tool rest 20 may be blade lengths, neck lengths, bag lengths, neck diameters, bag diameters calculated for tool type discrimination, tool data Lt; / RTI >

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation, Various changes and modifications will be possible.

10: tool 20: tool stand
100: tool type discrimination system 110: camera
120: Computer

Claims (9)

Photographing an image of a tool (hereinafter referred to as a " tool image ") mounted on the tool holder using a camera;
The computer receiving the tool image and extracting an outline image of the tool by detecting an edge of the tool;
Calculating a flute length, a neck length, a shank length, a neck diameter or a bag diameter of the tool from the contour image; And
And determining the type of the tool by inputting the blade length, the neck length, the bag length, the neck diameter, or the bag diameter into a data classifier.
The method according to claim 1,
Extracting the contour image,
Receiving the tool image and detecting a contour of the tool;
Confirming whether the contour is a closed curve;
Extracting an image of the contour line as the contour line image if the contour line is a closed curve; And
If the contour is not a closed curve, photographing conditions are changed to shoot the contour image again, and the contour of the tool is detected from the re-photographed tool image to be merged with the contour of the previously detected tool to determine whether the contour is a closed curve And returning to the step of identifying the tool type.
3. The method of claim 2,
The tool image may be an image photographed so that the blade of the tool faces downward, or an image of the tool being rotated in a state in which the blade of the tool is laterally directed,
Calculating the blade length, the neck length, the bag length, the neck diameter, or the bag diameter:
Detecting a bottom pixel in the contour image;
Calculating a width of the outline image from the height of the bottom pixel (hereinafter, referred to as 'first height') to the top;
Detecting a height (hereinafter, referred to as " second height ") at which the width change rate of the outline image is less than or equal to a first threshold change rate;
Detecting a height (hereinafter, referred to as a 'third height') at which the width change rate of the outline image becomes equal to or greater than a second threshold change rate after the second height detection; And
Detecting an uppermost pixel of the contour image,
The blade length is calculated as the distance between the first height and the second height,
The neck length is calculated as the distance between the second height and the third height,
The bag length is calculated as a distance between the third height and a height of the uppermost pixel (hereinafter referred to as a fourth height)
The neck diameter is calculated as the width of the contour image on the second height,
Wherein the bag diameter is calculated as the width of the contour image on the third height.
The method of claim 3,
Wherein the data classifier is a neural network classifier.
5. The method of claim 4,
When the type of the tool is determined in the neural network classifier, data relating to the determined blade length, neck length, bag length, neck diameter or bag diameter of the tool is transmitted to the tool, and when the tool is not discriminated, And returning to the step of determining the type of tool.
A computer program stored in a medium for performing a tool type discrimination method according to any one of claims 1 to 4 by functioning as a computer.
A storage medium capable of storing the computer program of claim 6; And
And a communication medium capable of transmitting the computer program stored in the storage medium through a communication network.
A camera for photographing an image of a tool installed on the tool stand; And
And a computer for installing the computer program of claim 6 and receiving a tool image from the camera and determining a tool type.
Bed;
A tool base provided on the bed and capable of machining a workpiece by mounting a tool; And
And the tool type discrimination system according to claim 8, wherein the tool type discrimination system discriminates the tool type by photographing the tool of the tool type.

Images