CN111638214B - Trademark consistency detection system and method based on machine vision - Google Patents

Abstract

The present invention proposes a trademark consistency detection system based on machine vision, including: including X/Z sliding table device, correction device, vision device and electric control device, X/Z slide table device, correction device, vision device are all connected with electric control device The devices are connected, and the X/Z sliding table device is used to install the correction device and carry the target forward at a constant speed; the correction device is used to correct the position of the target; the vision device includes an imaging module, an image processing module and a visual detection module, and the imaging module For the trademark image of the target object, the image processing module processes the image, and the visual detection module recognizes, detects and judges the image. The present invention also provides a detection method using the above system. The invention can realize 24-hour continuous operation, 100% read the information in energy efficiency stickers and barcodes, and can add different refrigerator models, which greatly improves the detection efficiency; and the later maintenance cost is small and easy to trace.

Description

Machine vision-based trademark consistency detection system and detection method

technical field

The invention relates to the technical field of machine vision recognition, in particular to a machine vision-based trademark consistency detection system, and also to a machine vision-based trademark consistency detection method.

Background technique

During the production process of refrigerators in the factory, a model barcode, a QR code for energy efficiency stickers, and a product logo need to be pasted on the front of the refrigerator, and a barcode that corresponds to the previous barcode needs to be pasted on the back of the refrigerator, so the refrigerator will be tested during the production process. Whether the barcodes on the front and back sides, the QR code on the energy efficiency sticker, and the logo are completely consistent. The traditional production process uses manual inspection. The inspection method is to arrange two inspectors on both sides of the assembly line to compare the barcode, energy efficiency sticker QR code, and Logo one by one against the template. However, this inspection method has many disadvantages. First, the refrigerator is always in motion on the assembly line, which is inconvenient for manual inspection and the detection accuracy is low; second, the inspection items include barcodes and QR codes, and manual inspection cannot directly read the information in the codes. The accuracy cannot be guaranteed; third, the results of manual testing are not easy to trace, and it is not convenient for later review.

Contents of the invention

The invention proposes a machine vision-based trademark consistency detection system and detection method, which solves the problems of low manual detection accuracy and difficult traceability in the prior art.

Technical scheme of the present invention is realized like this:

The trademark consistency detection system based on machine vision, including X/Z slide device, correction device, vision device and electric control device, X/Z slide device, correction device, vision device are all connected with the electric control device,

The X/Z sliding table device is used to install the correction device and carry the target object forward at a constant speed;

The correction device is used to correct the position of the target;

The visual device includes an imaging module, an image processing module and a visual detection module. The imaging module includes a code reader and an industrial camera. The code reader reads the trademark information and sends a trigger signal. The industrial camera captures an image, and the image processing module processes the image After processing, the visual inspection module recognizes, detects, and judges the image.

Optionally, the calibration device includes a cylinder, a push handle, a solenoid valve, and a photoelectric sensor, wherein the photoelectric sensor includes a first photoelectric sensor and a second photoelectric sensor, and when the first photoelectric sensor detects a target, the optical signal is converted into an electrical signal After receiving the signal, the solenoid valve triggers the cylinder to work, and the push handle connected to one end of the cylinder corrects the position of the target. When the target passes the second photoelectric sensor, it triggers the code reader to read the barcode behind the target.

Optionally, the correction device further includes a baffle mounted on the X/Z-axis slide device to prevent accidents of the target object during transportation.

Optionally, the correction device further includes a positioning member installed on the X/Z-axis slide device to adjust the position of the target in time to avoid deviation of the target.

Optionally, the piston of the cylinder adopts a heterosexual bidirectional sealing structure, which is compact in size and runs smoothly.

Optionally, the solenoid valve adopts a spool type structure, which has good sealing performance, sensitive response, small frictional resistance, no need for oil lubrication, low starting air pressure and long service life.

Optionally, the X/Z-axis sliding table device includes an X/Z-axis bracket, a screw drive mechanism, a servo motor, an X-axis slide table and a Z-axis slide table, and a screw drive mechanism, a servo The motor, the X-axis slide table and the Z-axis slide table, the X-axis slide table is installed on the Z-axis slide table, and the X-axis slide table can slide on the Z-axis slide table, the screw drive mechanism includes a screw, a nut and a reverser , the reverser is installed on the nut, the screw is driven to rotate by the servo motor, the nut converts the rotary motion into linear motion, and drives the X-axis sliding table and the Z-axis sliding table to realize linear reciprocating motion.

Optionally, the X/Z-axis support is a "ten"-shaped support, including an X-axis support and a Z-axis support, and a manipulator of a vision device is installed on the X-axis support. Optionally, the "ten"-shaped bracket is made of aluminum alloy materials with different lengths and a cross-section of 100mm*100mm, which are fixed by screws.

Optionally, X-axis forward limiters and X-axis reverse limiters are respectively installed at both ends of the X-axis slide table.

Optionally, the Z-axis slide table is provided with a Z-axis forward limiter and a Z-axis reverse limiter.

Optionally, the electric control device includes an industrial computer, a filter, a relay, a servo motor, a servo driver, and a PLC. The second photoelectric sensor detects the target object, and the input signal is transmitted to the code reader after being processed by the filter, and the code is read. The reader enters the external continuous trigger state until it detects the last barcode and reads the barcode information, and the barcode reader stops triggering; after the background module detects the barcode information, it matches the information in the database and sends instructions to the PLC, and the servo driver follows the instructions of the PLC Execute the actions of starting, running, and resetting, and control the movement of the servo motor according to the X/Z axis parameters set by the industrial computer, and cooperate with the industrial camera to capture images.

Optionally, the electric control device further includes a circuit breaker.

Optionally, the imaging module further includes a light source.

Optionally, the light source is a bar-shaped LED light source, which provides uniform illumination of a large area, stable performance, and long life, and is suitable for instant surface illumination of larger-sized objects.

Optionally, the visual detection module includes barcode information, energy efficiency sticker information, Logo standard value unit and measured value unit, the standard value unit stores the image information or character information input into the database in advance, and the measured value unit stores the camera every time The captured image and the processed image after the imaging operation.

The present invention also provides a machine vision-based trademark consistency detection method, which is detected based on the above-mentioned system, including the following steps:

The calibration device corrects the position of the target;

The industrial camera takes an image of the trademark of the target object, and the image processing module processes the captured image, and then transmits the image to the visual inspection unit;

The visual detection unit recognizes, detects and judges the trademark in the image, and outputs the result parameters.

Optionally, the position correction includes the following steps: the target object on the X-axis slide table moves at a constant speed on the X-axis slide table, and after the first photoelectric sensor senses the target object, it converts the optical signal into an electrical signal and transmits it to the solenoid valve, and the solenoid valve After receiving the signal, the cylinder is triggered to work, and the push handle connected to one end of the cylinder corrects the position of the target, and unifies the position of the target to one side of the X-axis slide table.

Optionally, the industrial camera takes an image of the trademark of the target object, including the following steps: the second photoelectric sensor detects the target object, transmits an input signal to the code reader after filter processing, and the code reader enters an external continuous trigger state, until the barcode is detected and the barcode information is read, the barcode reader stops triggering; after the background module detects the barcode information, it matches the information in the database and sends instructions to the PLC, and after the servo drive receives the start instruction from the PLC, X / The Z-axis sliding table device starts to run, the servo motor controlling the Z-axis sliding table rotates forward, the manipulator drives the industrial camera to move upward, and when it reaches the Z-axis coordinate position of the energy efficiency sticker, the servo motor controlling the Z-axis sliding table stops working; The servo motor of the X-axis sliding table rotates forward, and the manipulator moves to the right. After reaching the X-axis coordinate of the energy efficiency sticker, the servo motor of the X-axis sliding table is controlled to stop working, and the industrial camera takes the image of the energy efficiency sticker and the barcode; the energy efficiency sticker is completed. After capturing the image with the barcode, the servo motor controlling the Z-axis sliding table starts to rotate in reverse, and the manipulator moves downward. After reaching the Z-axis coordinate position of the Logo, the servo motor controlling the Z-axis sliding table stops working; controlling the X-axis sliding table The servo motor rotates in the opposite direction, and the manipulator moves to the left. After reaching the X-axis coordinate of the Logo, the servo motor controlling the X-axis sliding table stops working. The industrial camera reaches the coordinate position of the Logo, the industrial camera takes the image of the Logo, and the manipulator resets. Return to the starting position.

Optionally, before the visual detection unit recognizes, detects and judges the trademark in the image, it also performs visual configuration, including the following steps:

Training Logo image;

Set the QR code of the energy efficiency sticker;

set barcode;

View reports.

Optionally, the training Logo image includes the following steps:

Select the model to train;

Grab the image;

Determine the area and center origin of image training;

Determine the training parameters and perform training.

Optionally, when training the Logo image, a single character training is performed on the Logo, the training area is a single character and the center origin is re-determined, which can improve the detection accuracy and avoid situations such as incomplete pasting of the Logo or part of characters falling off during the production process.

Optionally, said setting the two-dimensional code of the energy efficiency sticker includes the following steps:

Manually add energy efficiency sticker QR code information;

Determine the detection area and center origin.

Optionally, the setting barcode includes the following steps:

Select the training model;

Determine the area and center origin of image training;

to train.

The beneficial effects of the present invention are:

1. The present invention utilizes machine vision combined with a visual inspection system built with automatic control technology to realize barcodes on the front and rear of the refrigerator, two-dimensional codes for energy efficiency stickers, and logo consistency inspections. After being put into actual production, it was found that compared with manual inspection, the trademark consistency inspection system can realize 24-hour continuous operation, 100% read the information in energy efficiency stickers and barcodes, and can realize instant image detection without being affected by the assembly line, greatly improving Improve the detection efficiency; the present invention can accurately read the information in the code, and complete the comparison and matching in the detection system, which can ensure 100% detection accuracy; the detection results will be automatically stored, including image results and final detection results, It is convenient to find and trace the cause of the problem later.

2. The present invention can replace 2-3 workers to complete the detection work of the visual part, and the later maintenance cost is small, easy to trace, and new target models can be added at any time. Even if the assembly line is replaced, the visual detection system can still be used. In the long run, the application of machine vision inspection equipment can also save a large part of the production cost for manufacturers, with high cost performance, and has great application space and promotion value in the white goods industry.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is a structural schematic diagram of a trademark consistency detection system based on machine vision of the present invention;

Fig. 2 is the structural diagram of the X/Z axis sliding table device of the present invention;

Fig. 3 is a working schematic diagram of the calibration device of the present invention.

Reference signs:

1. Push handle, 2. Refrigerator, 3. Baffle, 4. First photoelectric sensor, 5. Industrial computer, 6. Second photoelectric sensor, 7. Code reader, 8. Light source, 9. Industrial camera, 10. Manipulator , 11. Display screen, 12. Positive position piece, 13. Z-axis positive limit piece, 14. X-axis reverse limit piece, 15. Z-axis reverse limit piece, 16. X-axis positive limit piece pieces.

Detailed ways

The following description and the accompanying drawings sufficiently illustrate specific embodiments herein to enable those skilled in the art to practice them. Portions and features of some embodiments may be included in or substituted for those of other embodiments. The scope of the embodiments herein includes the full scope of the claims, and all available equivalents of the claims. Herein, the terms "first", "second", etc. are only used to distinguish one element from another element without requiring or implying any actual relationship or order between these elements. In fact the first element can also be called the second element and vice versa.

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

As shown in Figures 1-3, this application discloses a machine vision-based trademark consistency detection system, including an X/Z slide device, a correction device, a vision device and an electric control device, an X/Z slide device, a calibration The device and visual device are connected with the electric control device,

The X/Z sliding table device is used to install the correction device and carry the target object forward at a constant speed;

The correction device is used to correct the position of the target;

The visual device includes an imaging module, an image processing module and a visual inspection module, the imaging module includes a code reader 7 and an industrial camera 9, the code reader 7 reads the trademark information and sends a trigger signal, the industrial camera 9 takes images, and the image The processing module processes the image, and the visual detection module recognizes, detects, and judges the image.

In some embodiments, the target object is a refrigerator, and in other embodiments, the target object may be other white goods.

Optionally, as shown in Figure 3, the calibration device includes a cylinder, a pusher 1, a solenoid valve, and a photoelectric sensor, wherein the photoelectric sensor includes a first photoelectric sensor 4 and a second photoelectric sensor 6, and the first photoelectric sensor 4 detects When the refrigerator is 2, the optical signal is converted into an electrical signal and transmitted to the solenoid valve. The solenoid valve triggers the cylinder to work after receiving the signal, and the push handle 1 connected to one end of the cylinder corrects the position of the refrigerator 2, and the refrigerator 2 is triggered when it passes the second photoelectric sensor 6. The code reader 7 reads the barcode on the back of the refrigerator 2.

Optionally, the correction device further includes a baffle 3 installed on the X/Z-axis slide device to prevent accidents of the target object during transportation.

Optionally, the correction device further includes a positioning member 12 installed on the X/Z-axis slide device to adjust the position of the target object in time to avoid deviation of the target object.

Optionally, the present invention adopts AirTAC SC series standard cylinders, including cylinder barrels, end caps, pistons, piston rods and seals, and circular metal devices that guide the pistons to perform linear reciprocating motions. The specific structure and working principle are in No need to go into details here, the piston adopts a heterosexual two-way sealing structure, which is compact in size and runs smoothly; one end of the cylinder is connected to the pusher 1, and under the guidance of the cylinder, the pusher 1 pushes the refrigerator 2 to one side of the X-axis slide table smoothly.

Optionally, the present invention adopts the AirTAC 4V300 series solenoid valve with a spool structure, good sealing performance, sensitive response, small frictional resistance, no need for lubrication, low starting air pressure, and long service life. There are two round holes on the solenoid valve to insert the air pipe, and the force generated by the induction of the solenoid coil drives the action of the valve core to realize the switch of the valve, the switching of the flow direction and the size of the airflow. Set the value of the airflow to be moderate. If the airflow is too large, the thrust of the cylinder will be greater, which may overturn the refrigerator 2; If the positions are not uniform, it will affect the imaging and detection of the visual device.

Optionally, the photoelectric sensor used in the present invention is a Panasonic small photoelectric sensor CX-400, which converts light into electrical signals. The photoelectric sensor is resistant to inverter circuits and has strong anti-interference ability. The front cover and display cover are made of non-corrodible polycarbonate material, which can be used without sensitivity adjustment; the two-hole design saves installation time; special optical lens design, manual The color of the parts is less affected; the infrared beam is dust and dirt resistant, and the detection distance can reach 15cm. The photoelectric sensor judges whether there is a refrigerator 2 passing through by sensing the change of the light signal. Therefore, the surface of the photoelectric sensor needs to be cleaned regularly to prevent the surface from being covered with too much dust and affecting the photosensitive components to receive the light signal. The photosensitive components in the photoelectric sensor are relatively sensitive, and the refrigerators 2 on the X-axis sliding table should be separated from each other by a certain distance, so as to avoid continuous triggering of the photoelectric sensors, resulting in changes in the position of the refrigerator 2 that has passed the calibration device.

Optionally, as shown in Figure 2, the X/Z-axis sliding table device includes an X/Z-axis bracket, a screw drive mechanism, a servo motor, an X-axis sliding table and a Z-axis sliding table, and the X/Z-axis bracket is equipped with Screw transmission mechanism, servo motor, X-axis slide table and Z-axis slide table. The X-axis slide table is installed on the Z-axis slide table, and the X-axis slide table can slide on the Z-axis slide table. The screw drive mechanism includes a screw , a nut and a reverser, the reverser is mounted on the nut, the screw is driven to rotate by a servo motor, the nut converts the rotary motion into a linear motion, and drives the X-axis sliding table and the Z-axis sliding table to realize linear reciprocating motion. When the X/Z-axis sliding table device is in the automatic running state, after receiving the instruction from the electronic control device, the servo motor will move freely according to the preset parameters to cooperate with the camera to collect images. The running speed of the manipulator 10 is 30mm/s .

Optionally, the X/Z-axis support is a "ten"-shaped support, including an X-axis support and a Z-axis support, and the manipulator 10 of the vision device is installed on the X-axis support. Optionally, the "ten"-shaped bracket is made of aluminum alloy materials with different lengths and a cross-section of 100mm*100mm, which are fixed by screws.

Optionally, X-axis forward limiters 16 and X-axis reverse limiters 14 are respectively installed at both ends of the X-axis slide table.

Optionally, a Z-axis forward limiter 13 and a Z-axis reverse limiter 15 are provided on the Z-axis slide table.

Optionally, the lead screw transmission mechanism includes a screw, a nut and a reverser, the screw is driven to rotate through the servo motor, and the rotational motion is converted into linear motion through the nut, which drives the mechanical structure connected to it to achieve linear motion and reciprocating motion, It has the advantages of high transmission efficiency and accurate positioning.

Optionally, the present invention adopts the Cognex DataMan262QL code reader, equipped with decoding algorithm 1DMax and Hotbars II image processing and analysis tools, 1DMax can read the most difficult barcodes, and it is also the most comprehensive one-dimensional barcode algorithm, which can accurately analyze UPC/EAN/JAN, Code39, Code128 and other internationally common one-dimensional barcodes, when the code type is known, directly check the option of this type of code in the software, which can greatly shorten the decoding time and improve the decoding efficiency; HotbarsⅡ Image processing analysis technology combines excellent signal fidelity and illumination speed to provide superior decoding performance for DataMan series barcode readers; and integrated illumination, with liquid lens technology, can automatically adjust the focus.

Optionally, for the code reader 7 and the DataMan module used in the present invention, the user can set the parameters of the code reader through the DataMan module. The model of the connected code reader, real-time image, various parameters and code reading results can be displayed in the DataMan module, and parameters such as tuning, testing, light source setting, code reading type, data format, and communication settings can also be set through the DataMan module .

Optionally, the imaging module further includes a light source 8 .

Optionally, the light source 8 is a bar-shaped LED light source, which provides uniform illumination of a large area, stable performance, and long life, and is suitable for instant surface illumination of larger-sized objects.

Optionally, the visual detection module includes barcode information, energy efficiency sticker information, Logo standard value unit and measured value unit, the standard value unit stores the image information or character information input into the database in advance, and the measured value unit stores the camera every time the image is taken The images collected after the job and the processed images.

Optionally, the industrial camera 9 is responsible for image collection of the two-dimensional codes, logos, and barcodes on the front of the refrigerator 2, and the images are processed, analyzed, identified, and detected by the image processing module and the visual detection module, and compared with the pre-trained template Compare and get the score, when the score is higher than or equal to the set threshold, it is judged as qualified, otherwise it is judged as unqualified; at the same time, the information or images collected by the code reader 7 and the industrial camera 9 are displayed on the display screen 11 , output the detection results and record the detection results.

The electric control device includes an industrial computer 5, a filter, a relay, a servo motor, a servo driver and a PLC, and the second photoelectric sensor 6 detects the refrigerator 2, and transmits an input signal to the code reader 7 after being processed by the filter, and the code reader 7 Enter the external continuous trigger state until the barcode is detected and the barcode information is read, the barcode reader 7 stops triggering; after the background module detects the barcode information, it matches the information in the database and sends instructions to the PLC, and the servo driver follows the PLC. The command executes the actions of starting, running, and resetting, and controls the movement of the servo motor according to the X/Z axis parameters set by the industrial computer 5, and cooperates with the industrial camera 9 to take images.

Optionally, the electric control device further includes a circuit breaker.

The present invention also provides a machine vision-based trademark consistency detection method, which is detected based on the above-mentioned system, including the following steps:

The correction device corrects the position of the refrigerator 2;

The industrial camera 9 captures an image of the trademark of the refrigerator 2, and the image processing module processes the captured image, and then transmits the image to the visual detection unit;

The visual detection unit recognizes, detects and judges the trademark in the image, and outputs the result parameters.

Optionally, the position correction includes the following steps: the refrigerator 2 moves at a constant speed from left to right on the X-axis slide table, after the first photoelectric sensor 4 senses the refrigerator 2, converts the optical signal into an electrical signal and transmits it to the solenoid valve, The solenoid valve triggers the cylinder to work after receiving the signal, and the push handle 1 connected to one end of the cylinder corrects the position of the refrigerator 2 to unify the position of the refrigerator 2 to one side of the X-axis slide table.

Optionally, the industrial camera takes an image of the trademark of the refrigerator 2, including the following steps: the second photoelectric sensor 6 detects the refrigerator 2, transmits the input signal to the code reader 7 after filter processing, and the code reader 7 enters the external continuous In the trigger state, until the barcode is detected and the barcode information is read, the barcode reader 7 stops triggering; after the background program detects the barcode information, it matches the information in the database and sends instructions to the PLC, and the servo driver receives the start instruction from the PLC , the X/Z-axis sliding table device starts to run, the servo motor controlling the Z-axis rotates forward, the manipulator drives the camera to move upward, and when it reaches the Z-axis coordinate position of the energy efficiency sticker, the servo motor controlling the Z-axis stops working; the servo motor controlling the X-axis The servo motor rotates forward, and the manipulator moves to the right. After reaching the X-axis coordinate of the energy efficiency sticker, the servo motor controlling the X axis stops working, and the camera captures the image of the energy efficiency sticker and the barcode; after the energy efficiency sticker and the barcode are captured, the control The servo motor on the Z-axis starts to rotate in reverse, and the manipulator 10 moves downward. After reaching the Z-axis coordinate position of the Logo, the servo motor that controls the Z-axis stops working; the servo motor that controls the X-axis rotates in the opposite direction, and the manipulator 10 moves to the left. After arriving at the X-axis coordinate of the Logo, the servo motor controlling the X-axis stops working, the camera arrives at the coordinate position of the Logo, the camera takes an image of the Logo, the manipulator 10 resets, and returns to the starting position.

Optionally, before the visual detection unit recognizes, detects and judges the trademark in the image, it also performs visual configuration, including the following steps:

Training Logo image;

Set the QR code of the energy efficiency sticker;

set barcode;

View reports.

Optionally, the training Logo image includes the following steps:

Select the model to train;

Grab the image;

Determine the area and center origin of image training;

Determine the training parameters and perform training.

Optionally, when training the Logo image, a single character in the Logo is trained, the training area becomes a single character and the center origin is re-determined, which can improve the detection accuracy and avoid incomplete pasting of the Logo or part of the characters falling off during the production process, etc. Condition.

Optionally, said setting the two-dimensional code of the energy efficiency sticker includes the following steps:

Manually add energy efficiency sticker QR code information;

Determine the detection area and center origin.

Optionally, the setting barcode includes the following steps:

Select the training model;

Determine the area and center origin of image training;

to train.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of the present invention. within the scope of protection.

Claims (1)

1. A machine vision-based trademark consistency detection system, characterized in that it includes: X/Z sliding table device, correction device, visual device and electric control device, X/Z sliding table device, correction device, visual device are all connected with the electric control device, X/Z sliding table device is used to install the correction device and carry the target things move forward at a uniform speed; The correction device is used to correct the position of the target; The vision device includes an imaging module, an image processing module and a visual detection module. The imaging module includes a code reader and an industrial camera. The code reader reads the trademark information of the target and sends a trigger signal. The industrial camera captures an image and processes the image. The module processes the image, and the visual detection module recognizes, detects, and judges the image; The correction device includes a cylinder, a pusher, a solenoid valve and a photoelectric sensor, wherein the photoelectric sensor includes a first photoelectric sensor and a second photoelectric sensor, and when the first photoelectric sensor detects an object, the light signal is converted into an electrical signal and transmitted to the solenoid valve , the solenoid valve triggers the cylinder to work after receiving the signal, and the push handle connected to one end of the cylinder corrects the position of the target, and when the target passes the second photoelectric sensor, the code reader is triggered to read the barcode behind the target; The correction device also includes a baffle plate and a positioning member installed on the X/Z axis slide device; The X/Z-axis sliding table device includes an X/Z-axis support, a screw drive mechanism, a servo motor, an X-axis slide table and a Z-axis slide table, and a screw drive mechanism, a servo motor, and an X-axis are installed on the X/Z-axis support. The sliding table and the Z-axis sliding table, the X-axis sliding table is installed on the Z-axis sliding table, and the X-axis sliding table can slide on the Z-axis sliding table; the screw transmission mechanism includes a screw, a nut and a reverser, and the reverser The direction device is installed on the nut, and the screw is driven to rotate by the servo motor, and the nut converts the rotational motion into a linear motion to drive the X-axis sliding table and the Z-axis sliding table to realize linear reciprocating motion; the two ends of the X-axis sliding table are respectively equipped with X-axis A positive limiter and an X-axis reverse limiter, the Z-axis slide table is provided with a Z-axis forward limiter and a Z-axis reverse limiter; The visual detection module includes barcode information, energy efficiency sticker information, Logo standard value unit and actual measurement value unit. The standard value unit stores the image information or character information input into the database in advance, and the actual measurement value unit stores the image information after the camera completes the imaging operation each time. The collected image and the processed image; Detection based on the above system includes the following steps: The calibration device corrects the position of the target; The industrial camera takes an image of the trademark of the target object, and the image processing module processes the captured image, and then transmits the image to the visual inspection unit; The visual detection unit recognizes, detects and judges the trademark in the image, and outputs the result parameters; The position correction includes the following steps: the target object moves at a constant speed on the X-axis slide table, the first photoelectric sensor senses the target object, converts the optical signal into an electrical signal and transmits it to the solenoid valve, and the solenoid valve triggers the cylinder to work after receiving the signal, The push handle connected to one end of the cylinder corrects the position of the target and unifies the position of the target to one side of the X-axis slide table; The industrial camera takes an image of the trademark of the target object, including the following steps: the second photoelectric sensor detects the target object, transmits an input signal to the code reader after being processed by a filter, and the code reader enters an external continuous trigger state until the detection Until the barcode is read and the barcode information is read, the barcode reader stops triggering; after the background module detects the barcode information, it matches the information in the database and sends instructions to the PLC. After the servo drive receives the start instruction from the PLC, the X/Z axis slides The stage device starts to run, the servo motor controlling the Z-axis sliding table rotates forward, and the manipulator drives the industrial camera to move upwards. After reaching the Z-axis coordinate position of the energy efficiency sticker, the servo motor controlling the Z-axis sliding table stops working; controls the X-axis sliding table The servo motor rotates forward, and the manipulator moves to the right. After reaching the X-axis coordinate of the energy efficiency sticker, the servo motor controlling the X-axis sliding table stops working, and the industrial camera captures the image of the energy efficiency sticker and the barcode; completes the capture of the energy efficiency sticker and the barcode After the imaging, the servo motor controlling the Z-axis sliding table starts to rotate in reverse, and the manipulator moves downward. After reaching the Z-axis coordinate position of the Logo, the servo motor controlling the Z-axis sliding table stops working; the servo motor controlling the X-axis sliding table reverses. Rotate to the left, the manipulator moves to the left, and after reaching the X-axis coordinate of the Logo, the servo motor controlling the X-axis sliding table stops working, the industrial camera reaches the coordinate position of the Logo, the industrial camera takes the image of the Logo, resets the manipulator, and returns to the starting point Location; Before the visual detection unit recognizes, detects and judges the trademark in the image, it also performs visual configuration, including the following steps: Training Logo image; Set the QR code of the energy efficiency sticker; set barcode; View reports; Described training Logo image comprises the steps: Select the model to train; Grab the image; Determine the area and center origin of image training; Determine the training parameters and perform training; When training the Logo image, the Logo is trained on a single character, the training area is a single character and the center origin is re-determined.