TWI733577B - Real-time tool image identification system - Google Patents

Abstract

The utility model relates to a real-time tool image identification system, which comprises an IoT Router, a spindle camera, and a controller connected to the IoT Router separately, and the controller is installed in a machine tool. The spindle camera takes the knife holder and the tool image installed on the machine tool host as the spindle tool image, and there is a cloud platform connected with the IoT Router signal. It is used to receive the spindle tool image and the equipment tool data provided by the controller, and the real-time tool image identification system of the client platform analyzes and processes the spindle tool image and compares it with the obtained equipment tool data. if there is a discrepancy, the controller is controlled through the IoT Router to make the machine tool send an alarm and stop, avoid using wrong or inappropriate tools during machining, and improve the quality of processing. Avoid tool and equipment failure.

Description

Tool real-time image recognition system

The invention relates to an image recognition system, in particular to a tool real-time image recognition system for performing tool image recognition through cloud applications.

The existing machine tool recognizes the tools in the tool magazine. The main method is to place the wireless sensor or barcode on the tool holder in advance, and then use the reader to read the identification code of the tool holder, and use the identification code to obtain the corresponding tool holder and the tool from the database. Relevant information, and then judge whether the handle and its tools in the pocket are correct.

The above-mentioned existing method of identifying the tool holder and the tool in the tool magazine only compares the label of the tool holder with the setting of the tool holder in the machining program, but this method cannot detect whether the tool in the tool holder is installed correctly, nor can it detect the wear of the tool in the tool holder. Whether the condition is suitable for machining, it is impossible to judge the wear of the tool or the condition of the tool being installed incorrectly, and the condition of using inappropriate or wrong tools during processing cannot be ruled out. It is easy to cause the failure of the machine tool or the damage of the machined parts.

Since the tool recognition method of the existing tool magazine cannot distinguish the condition of tool installation and wear, it will still use inappropriate or wrong tools for processing. For this reason, the present invention utilizes the Internet of Things system to cooperate with cloud service applications for tool identification, achieves the effect of tool and processing quality control, and reduces equipment failures caused by manual tool assembly errors or excessive tool wear.

In order to achieve the above-mentioned creative purpose, the present invention provides a real-time image recognition system for cutting tools, including:

An Internet of Things system includes an Internet of Things router, a spindle camera and a controller electrically connected to the Internet of Things router, the controller is installed in a machine tool, the machine tool is provided with a spindle, and the spindle camera photographs the spindle The image of the tool holder and the tool installed on it is the image of the spindle tool, the image of the spindle tool is externally transmitted by the IoT router, and the controller transmits the device tool data of the machine tool from the IoT router; and

A cloud platform, the cloud platform is connected with the IoT router signal to receive the spindle tool image and the equipment tool data, the cloud platform is provided with a database and a real-time tool image recognition system, the equipment tool data is stored in the database , The real-time tool image recognition system is connected with the database to obtain the equipment tool data of the machine tool. The real-time tool image recognition system analyzes and processes the spindle tool image and compares it with the equipment tool data. The Internet of Things router controls the controller, causing the machine tool to sound an alarm and shut down.

Further, in the present invention, when the real-time tool image recognition system analyzes and processes the spindle tool image and does not match the tool data of the equipment, the real-time tool image recognition system notifies the user of the machine tool via email or communication software of an alarm message And management staff.

Furthermore, the machine tool of the present invention is provided with a tool magazine, the tool magazine is electrically connected to the controller and controlled by the controller, the tool magazine has a plurality of pockets, and a handle and a tool are installed in each pocket; A first tool magazine camera and a second tool magazine camera are respectively electrically connected to the Internet of Things router. The first tool magazine camera and the second tool magazine camera shoot images of the tool holder and the tool of each pocket at different angles as the tool magazine. Tool image, the tool library tool image is transmitted by the Internet of Things router; the cloud platform receives the tool library tool image, the real-time tool image recognition system analyzes and processes the tool library tool image and compares it with the equipment tool data. If the comparison does not match, the controller will be controlled through the Internet of Things router to cause the machine tool to sound an alarm and shut down.

Preferably, in the present invention, when the real-time tool image recognition system analyzes and processes the tool image of the tool magazine and does not match the tool data of the equipment, the real-time tool image recognition system notifies the machine tool of an alarm message via email or communication software Users and management personnel.

When the present invention is used, the spindle camera of the Internet of Things system is used to shoot the spindle tool image, and the spindle tool image and the equipment tool data obtained by the controller are transmitted to the cloud platform through the Internet of Things router, and the real-time tool image of the cloud platform is used for identification The system analyzes and processes the images, and obtains realistic information such as tool number, tool appearance, or tool length, then compares the actual tool information with the corresponding data in the equipment tool data to confirm whether the tool in the tool holder is installed correctly and can be detected Whether the wear condition of the tool is suitable for processing, to achieve the effect of tool and processing quality control, and to reduce equipment failures caused by manual tool assembly errors or excessive tool wear.

In order to understand the technical features and practical effects of the present invention in detail, and implement it in accordance with the content of the specification, the preferred embodiments shown in the drawings are further described in detail as follows.

As shown in the preferred embodiment shown in FIGS. 1 to 3, the present invention provides a tool real-time image recognition system, which includes an Internet of Things system 10, a cloud platform 20 signally connected to the Internet of Things system 10, and a tool that provides the thing A machine tool 30 installed in the networked system 10, the machine tool 30 is provided with a controller 31, a spindle 32 and a tool magazine 33 electrically connected to the controller 31 and controlled by the controller 31, and a tool holder is installed on the spindle 32 And tools, the tool magazine 33 has a plurality of pockets 331, and a handle and a tool are installed in each pocket 331, of which:

The Internet of Things system 10 includes an Internet of Things router 11, a spindle camera 12, a first tool magazine camera 13, and a second tool magazine camera 14 electrically connected to the Internet of Things router 11, and the controller of the machine tool 30 31 is also electrically connected to the Internet of Things router 11; the spindle camera 12 shoots the image of the tool holder and the tool installed on the spindle 32 as a spindle tool image, and the spindle tool image is externally transmitted by the Internet of Things router 11; the first tool magazine camera 13 and the second tool magazine camera 14 respectively shoot images of the tool holder and tool of each tool pocket 331 from different angles as the tool image of the tool magazine. The image of the tool magazine is transmitted to the outside by the IoT router 11; the controller 31 transfers the tool to the outside. The equipment tool data of the machine 30 is transmitted to the outside by the Internet of Things router 11. The equipment tool data includes equipment processing status, cutting time and usage times of each tool, and current processing tool detection parameters such as tool length and tool compensation table.

The cloud platform 20 and the Internet of Things router 11 are wirelessly connected through 5G and other mobile communications, or can be connected in a wired manner through an Ethernet port, so as to receive the spindle tool image, the tool magazine image, and For the equipment tool data, the cloud platform 20 is provided with a database 21 and a real-time tool image recognition system 22. The equipment tool data is stored in the database 21. The real-time tool image recognition system 22 is connected to the database 21 to obtain the The equipment tool data of the machine tool 30; when the user runs the machine tool 30, the real-time tool image recognition system 22 will analyze and process the tool images of the tool magazine in real time, record the images of the tools used in the tool library 33, and then The tool number, and then the image and number of each tool are stored in the database 21 of the cloud platform 20, which is continuously updated to provide the construction and use of subsequent processing history.

When the machine tool 30 is operating, the real-time tool image recognition system 22 analyzes and processes the spindle tool image and the tool magazine tool image in a real-time manner, and confirms the tool number of each tool on the spindle 32 and the tool magazine 33. Tool appearance identification and tool length measurement, and the analysis and processing results are compared with the corresponding equipment tool data. If the comparison does not match, it means that the equipment tool data is wrong. At this time, the controller 31 will be controlled through the Internet of Things router 11. The machine tool 30 is caused to issue an alarm and shut down, and the user and management personnel of the machine tool are notified of the alarm message and the information of the equipment and tool found to have errors through e-mail or communication software.

Please refer to the flowchart shown in FIG. 2, which is the real-time recognition process of the real-time tool image recognition system 22 taking the spindle 32 as an example, and the steps are as follows:

(S01) Start: start the real-time identification process;

(S02) Obtain equipment tool data: obtain equipment tool data;

(S03) Obtain spindle tool image: acquire spindle tool image;

(S04) Image analysis and processing: image analysis and processing of the spindle tool;

(S05) Compare equipment tool data: compare the results of spindle tool image analysis and processing with equipment tool data;

(S06) Whether it matches: If the comparison results match, return to the step of obtaining equipment tool data and cycle again, if not match, proceed to the step of writing the machine tool alarm message;

(S07) Write machine tool alarm message: write an alarm signal to the machine tool 30 through the controller 31 to make the machine tool 30 sound an alarm and stop; and

(S08) Send alarm message: then notify the relevant personnel of the alarm message via e-mail or communication software.

In addition to the aforementioned real-time identification process, the present invention can also perform a tool magazine health check before the machine tool 30 is processed, and run the tool magazine operation program before the machine tool 30 is produced, so that the system can check the tools installed in the tool magazine 33 The tool performs image recording, and throws the acquired image of the tool magazine and the tool number to the cloud platform 20 for identification and data recording. When the application server of the cloud platform 20, that is, the real-time tool image identification system 22 receives When the tool library is imaged, the following steps shown in the flowchart shown in Figure 3 will be performed:

(S11) Start: Start the tool magazine health inspection process;

(S12) Upload the diagnostic numerical control program: upload the diagnostic numerical control program to the controller 31 of the machine tool 30;

(S13) Obtain equipment tool data: obtain the equipment tool data from the controller 31;

(S14) Execution program: the controller 31 of the machine tool 30 executes the diagnostic numerical control program to control the operation of the tool magazine 33 so that the first tool magazine camera 13 and the second tool magazine camera 14 take pictures of each tool with a known number ；

(S15) Obtain the processing status of the equipment: the processing status of the machine tool 30 is acquired by the controller 31;

(S16) Whether the program is over: then judge whether the program is over, if it is over, proceed to the step of diagnosis end, if not, go to the step of obtaining the current tool number;

(S17) Diagnosis ends: the tool magazine health check process ends;

(S21) Acquire the current tool number: the real-time tool image recognition system 22 obtains the number of the currently photographed tool from the Internet of Things system 10;

(S22) Obtain the tool image of the tool magazine: the real-time tool image identification system 22 obtains the image of the tool of the tool magazine currently shot from the Internet of Things system 10;

(S23) Image analysis and processing: Analyze and process the image of the tool in the tool magazine;

(S24) Compare equipment tool data: compare the results of the analysis and processing of the tool library tool image with the equipment tool data, perform tool number confirmation, tool appearance recognition and tool length measurement, and use the Internet of Things system 10 Obtain the tool compensation table in the machine tool 30, and compare whether there is a parameter setting error;

(S25) Whether it is normal: If the comparison is normal, return to the step of obtaining the processing state of the equipment, if it is abnormal, proceed to the step of writing the machine tool alarm message;

(S26) Write the machine tool alarm message: write an alarm signal to the machine tool 30 through the controller 31 to make the machine tool 30 sound an alarm and stop; and

(S27) Send alarm message: then notify relevant personnel of the alarm message via e-mail or communication software.

The foregoing descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of rights claimed by the present invention. All other equivalent changes or modifications completed without departing from the spirit disclosed by the present invention should be included in the present invention. Within the scope of patent application.

10: IoT system 11: IoT router 12: Spindle camera 13: The first tool magazine camera 14: The second tool magazine camera 20: Cloud platform 21: Database 22: Real-time tool image recognition system 30: machine tool 31: Controller 32: Spindle 33: tool magazine 331: Sword Hole S01-S08: Step S11-S17: steps S21-S27: steps

Figure 1 is a block diagram of a preferred embodiment of the present invention. Fig. 2 is a flow chart of real-time identification of a preferred embodiment of the present invention. Fig. 3 is a flowchart of a tool magazine health check according to a preferred embodiment of the present invention.

10: IoT system

11: IoT router

12: Spindle camera

13: The first tool magazine camera

14: The second tool magazine camera

20: Cloud platform

21: Database

22: Real-time tool image recognition system

30: machine tool

31: Controller

32: Spindle

33: tool magazine

331: Sword Hole

Claims (4)

A tool real-time image recognition system, including: An Internet of Things system includes an Internet of Things router, a spindle camera and a controller electrically connected to the Internet of Things router, the controller is installed in a machine tool, the machine tool is provided with a spindle, and the spindle camera photographs the spindle The image of the tool holder and the tool installed on it is the image of the spindle tool, the image of the spindle tool is externally transmitted by the IoT router, and the controller transmits the device tool data of the machine tool from the IoT router; and A cloud platform, the cloud platform is connected with the Internet of Things router signal to receive the spindle tool image and the equipment tool data, the cloud platform is provided with a database and a real-time tool image recognition system, the equipment tool data is stored in the database , The real-time tool image recognition system is connected with the database to obtain the equipment tool data of the machine tool. The real-time tool image recognition system analyzes and processes the spindle tool image and compares it with the equipment tool data. The Internet of Things router controls the controller, causing the machine tool to sound an alarm and shut down. The tool real-time image recognition system according to claim 1, wherein when the real-time tool image recognition system analyzes and processes the spindle tool image and does not match the tool data of the equipment, the real-time tool image recognition system uses email or communication software Notify the user and management personnel of the machine tool of the alarm message. The tool real-time image recognition system according to claim 1 or 2, wherein the machine tool is provided with a tool magazine, the tool magazine is electrically connected to and controlled by the controller, and the tool magazine has a plurality of tool pockets. Each knife pocket is equipped with a knife handle and a knife; a first tool magazine camera and a second tool magazine camera are respectively electrically connected to the Internet of Things router, and the first tool magazine camera and the second tool magazine camera shoot at different angles. The image of the handle and the tool of each pocket is the image of the tool in the tool magazine, which is transmitted by the IoT router; the cloud platform receives the image of the tool in the tool library, and the real-time tool image recognition system analyzes and processes the tool in the tool magazine The image is compared with the tool data of the equipment. If the comparison does not match, the controller will be controlled through the Internet of Things router to cause the machine tool to sound an alarm and stop. The tool real-time image recognition system according to claim 3, wherein when the real-time tool image recognition system analyzes and processes the tool image of the tool magazine and does not match the tool data of the equipment, the real-time tool image recognition system uses email or communication The software notifies the user and management personnel of the machine tool of the alarm message.

Images