Automobile standard part counting and sorting method and system based on AI automatic identification

Abstract

The application provides a kind of automobile standard piece counting and sorting method and system based on AI automatic identification, which comprises the following steps: establishing a feature database for storing automobile standard pieces;Obtain the features of the standard pieces to be sorted and match them with the database standard features to obtain a comprehensive recognition matching degree;According to the matching degree, control the opening and closing of the access door body on the feeding path;Through the magnetic type rotating sorting mechanism, the standard pieces are adsorbed and followed by the transport standard pieces, and the number of passes is limited by the guide limiting structure, so that they enter the falling guide channel in single piece state;Use the detection sensor to detect the pass signal and generate real-time cumulative sorting quantity;When the real-time cumulative quantity reaches the preset sorting quantity, control the magnetic type rotating sorting mechanism to stop running, and complete the quantitative counting and sorting. This scheme realizes closed-loop operation of identification screening, automatic feeding, single piece sorting, accurate counting and quantitative shutdown, improves efficiency, reduces the risk of wrong feeding, missing counting and misloading, and improves scene adaptability and traceability.

Description

Technical Field

[0001] This invention relates to the field of automotive standard parts production technology, and in particular to an AI-based automatic identification method and system for counting and sorting automotive standard parts. Background Technology

[0002] In the assembly and production processes of automotive automatic transmissions and engine flexible discs, the sorting, counting, and retrieval of standard automotive parts such as bolts and nuts have long relied on manual operations. Existing technologies and operating methods have many shortcomings and are difficult to meet the high-efficiency, precise, and standardized requirements of automotive production. Specific problems are as follows: Low sorting efficiency: Standard parts are picked up and counted manually one by one, with a single operation time of 30 seconds per unit. This can easily lead to production line blockage due to insufficient efficiency, and cannot meet the needs of automobile production lines for efficient counting and feeding of standard parts in mass production. Significant quality risks: The appearance differences of similar automotive standard parts are minimal, making manual visual identification difficult and prone to mis-dispensing and mis-assembly, with a mis-assembly rate of 3‰; moreover, after the torque converter parts are assembled and covered, subsequent processes cannot verify the assembly status, which can easily lead to situations where bolts are missing or not tightened to the specified torque, resulting in bolts loosening when the gearbox rotates at high speed and causing serious quality accidents such as gearbox damage. Poor traceability: Manual operations lack accurate counting and sorting records, making it difficult to quickly locate the problem link after quality issues occur, resulting in high traceability costs. Summary of the Invention

[0003] In view of the shortcomings of the existing technology, the purpose of this invention is to provide an AI-based automatic identification method for counting and sorting standard automotive parts, aiming to solve the technical problems of low sorting efficiency and low accuracy in the existing technology.

[0004] To achieve the above objectives, in a first aspect, the present invention provides: a method for counting and sorting automotive standard parts based on AI automatic identification, comprising the following steps: Establish a database of automotive standard parts features, which stores the appearance features, dimensional features, and thread specifications of automotive standard parts to be sorted. Image information of automotive standard parts to be sorted is obtained, and the identification features of automotive standard parts to be sorted are extracted based on the image information. The identification features are then matched with standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts. The opening and closing state of the access gate set on the feeding path is controlled according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches the preset matching threshold, the access gate is controlled to open, so that the standard parts of the automobile to be sorted can enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, the access gate is controlled to close. The standard automotive parts passing through the entry gate are transported to the magnetic rotary sorting mechanism, which uses the magnetic rotary sorting mechanism to attract and transport the standard automotive parts. When the standard automotive parts move to the guide limit position with the magnetic rotary sorting mechanism, the number of standard automotive parts passing through is limited by the guide limit structure, so that the standard automotive parts enter the falling guide channel in a single piece state. The detection sensor installed at the falling guide channel detects the passing signal of the automotive standard parts, and generates a real-time cumulative sorting quantity based on the passing signal; The real-time cumulative sorting quantity is compared with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, the magnetic rotary sorting mechanism is controlled to stop running to complete a quantitative counting and sorting of automotive standard parts.

[0005] According to one aspect of the above technical solution, the calculation expression for the comprehensive identification matching degree of the standard parts is as follows: ; ; In the formula, S represents the comprehensive identification matching degree of the standard part, and n represents the total dimension of the identification features. Let be the weight coefficient of the i-th identification feature. Let be the matching coefficient of the i-th identification feature.

[0006] According to one aspect of the above technical solution, the magnetic rotary sorting mechanism includes a magnetic rotary disk driven by a drive motor. The back of the magnetic rotating disk is provided with a magnetic structure, so that the automotive standard parts entering the magnetic rotating sorting mechanism are attracted to the magnetic rotating disk and rotate with the magnetic rotating disk to the guide limiting structure.

[0007] According to one aspect of the above technical solution, when the automotive standard part rotates to the guide limiting structure, the guide limiting structure constrains the posture and passing position of the automotive standard part, so that only one automotive standard part is allowed to pass at the same time. When the standard automotive component continues to rotate to the entrance position of the falling guide channel, the standard automotive component is released from the magnetic attraction of the magnetic structure and enters the falling guide channel under the action of gravity.

[0008] According to one aspect of the above technical solution, the detection sensor is an infrared sensor; When the standard automotive component passes through the falling guide channel, the infrared sensor is triggered to generate a passing signal; The counting controller updates the real-time cumulative sorting quantity according to the signal, and controls the drive motor to keep running when the real-time cumulative sorting quantity is less than the preset sorting quantity, and controls the drive motor to stop running when the real-time cumulative sorting quantity is equal to the preset sorting quantity.

[0009] According to one aspect of the above technical solution, the preset sorting quantity is set by a counting controller, and the preset sorting quantity is a positive integer between 1 and 100; Once a quantitative counting and sorting cycle is completed, in response to the part-picking signal generated by the pressing down of the receiving tray, the magnetic rotary sorting mechanism is restarted to enter the next cycle of automotive standard parts counting and sorting.

[0010] According to one aspect of the above technical solution, the method further includes: Select the power supply mode, which includes AC power supply mode and DC power supply mode; The input voltage is converted to the device's rated operating voltage via a power conversion module and a voltage regulation module. After power is turned on, the image acquisition unit, drive motor and counting unit perform self-tests; When all inspection units are in normal condition, a workable prompt signal is output; When any self-test unit is in an abnormal state, it outputs a fault warning signal and restricts the start of sorting operations.

[0011] On the other hand, this solution also provides an AI-based automatic identification system for counting and sorting standard automotive parts, including: The data module is used to establish a database of automotive standard parts features, which stores the appearance features, size features, and thread specifications of automotive standard parts to be sorted. The feature matching module is used to acquire image information of automotive standard parts to be sorted, extract the identification features of automotive standard parts to be sorted based on the image information, and match the identification features with the standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts. The access module is used to control the opening and closing state of the access gate set on the feeding path according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches the preset matching threshold, the access gate is controlled to open, allowing the automotive standard parts to be sorted to enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, the access gate is controlled to close. The conveying module is used to convey the standard automotive parts that pass through the entry gate to the magnetic rotary sorting mechanism, which uses the magnetic rotary sorting mechanism to attract and follow the standard automotive parts. The limiting module is used to limit the number of automotive standard parts passing through when the automotive standard parts move to the guide limiting position with the magnetic rotary sorting mechanism, so that the automotive standard parts enter the falling guide channel in a single piece state. The counting module is used to detect the passing signal of the automotive standard parts through the detection sensor set at the falling guide channel, and generate a real-time cumulative sorting quantity based on the passing signal; The quantitative module is used to compare the real-time cumulative sorting quantity with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, the magnetic rotary sorting mechanism is controlled to stop running to complete the quantitative counting and sorting of automotive standard parts.

[0012] Compared with existing technologies, the advantages of this invention are as follows: By combining AI image recognition, electromagnetic access control, magnetic rotary sorting, guided and limited single-piece conveying, and infrared sensing counting, a closed-loop automated operation of automotive standard parts is achieved, from identification and screening, automatic feeding, single-piece sorting, accurate counting, to quantitative shutdown. Compared with manual sorting and counting, this invention can effectively improve the sorting and counting efficiency of automotive standard parts, reduce the problems of mis-dispensing, omissions, under-counting, and over-counting caused by manual identification and counting, and intercept non-target standard parts or abnormal standard parts from the source through identification matching degree judgment and abnormal alarm, reducing the risk of mis-assembly and omissions in the automotive assembly process. At the same time, by dynamically comparing the preset sorting quantity with the real-time cumulative quantity, flexible quantitative supply of standard parts required for different assembly processes can be achieved. With the AC / DC multi-source power supply structure, the adaptability and ease of use of the device in automotive assembly workshops, repair stations, and temporary operation scenarios are improved, thereby improving the accuracy, continuity, and traceability of automotive standard parts sorting operations. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating the AI-based automatic identification method for counting and sorting automotive standard parts in the first embodiment of the present invention. Figure 2 This is a structural block diagram of the automotive standard parts counting and sorting system based on AI automatic recognition in the second embodiment of the present invention; The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0014] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0015] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0017] Example 1 This embodiment provides an AI-based automatic recognition method for counting and sorting automotive standard parts. This method can be applied to operational scenarios requiring quantitative supply of bolts and nuts for automotive standard parts, such as the assembly of hydraulic torque converters and engine flexible discs, and the assembly of sealed components. In this embodiment, the automotive standard parts counting and sorting device implementing this method includes a frame, an image recognition unit, an access gate, a magnetic rotary sorting mechanism, a guide and limit structure, a drop guide channel, a detection sensor, a counting controller, a receiving tray, limit switches, an alarm unit, and a multi-source power supply unit. The image recognition unit includes an AI camera and an image recognition module; the access gate is preferably an electromagnetically controlled gate. The magnetic rotary sorting mechanism includes a drive motor, a coupling, bearings, and a magnetic rotary disk. The magnetic rotary disk is mounted on the frame via bearings, and the drive motor is connected to the magnetic rotary disk via a coupling. A permanent magnet or magnetic attractor is provided on the back of the magnetic rotary disk to attract and transport the automotive standard parts. The falling guide channel is positioned corresponding to the tangential direction of the top edge of the magnetic rotating disk, with an optimal tilt angle of 30°. The detection sensor is preferably an infrared sensor, embedded in the inner wall of the falling guide channel and forming a preset angle with its axis, preferably 15°. A limit switch is fixed to the bottom support structure of the receiving tray, used to generate a restart signal when the operator presses down on the receiving tray to retrieve the part. The multi-source power supply unit includes an AC power interface, a DC power interface, a power conversion module, and a voltage regulation module. The AC power interface is used to connect to a 220V AC power supply, and the DC power interface is used to connect to a 24V or 48V DC power supply. The power conversion module and voltage regulation module convert different input voltages into the operating voltage required by the image recognition unit, drive motor, detection sensor, and counting controller.

[0018] Please see Figure 1The figure shows a flowchart of the AI-based automatic identification method for counting and sorting automotive standard parts in the first embodiment of the present invention. As shown in the figure, the method includes the following steps: Step S100: Establish an automotive standard parts feature database, which stores the appearance features, size features, and thread specifications of the automotive standard parts to be sorted.

[0019] Specifically, in this embodiment, the standard features of the automotive standard parts to be sorted are pre-entered into the image recognition module to establish an automotive standard parts feature database. These standard features include appearance features, dimensional features, and thread specification features. They may further include one or more of the following: end face shape, head contour, thread profile, shank length, diameter, color grayscale, edge contour, surface defect features, and structural integrity features. For example, the appearance, length parameters, and thread specification parameters of commonly used gearbox assembly bolts such as M8×20mm and M10×25mm can be entered into the automotive standard parts feature database for subsequent identification and comparison of the actual automotive standard parts being processed.

[0020] After establishing a database of automotive standard parts features, a preset sorting quantity for a single sorting operation is set via a counting controller. This preset sorting quantity can be flexibly set between 1 and 100 pieces according to the needs of the automotive assembly process. Before sorting operations, the power supply mode is selected based on the on-site production conditions. When in a fixed automotive assembly workshop, a 220V AC power supply can be connected via the AC power interface; when in mobile operations, repair stations, or temporary work scenarios, a 24V or 48V DC power supply can be connected via the DC power interface. The power conversion module and voltage regulation module convert the input voltage into the operating voltage required by the device, for example, converting the input voltage into the 24V operating voltage required for the drive motor, and providing 12V or 24V operating voltage for the image recognition unit, counting controller, and detection sensors.

[0021] After power is connected, the device enters a self-test state, and the AI ​​camera, drive motor, and counting system start up sequentially. The self-test time is preferably 3 seconds. When the image recognition unit, drive motor, counting controller, detection sensor, and power supply unit are all in normal condition, the indicator light will show green, and the device will enter the workable state. When any functional unit is in an abnormal state, the indicator light will show red or output a fault alarm signal, and the device will be restricted from entering the sorting and counting operation.

[0022] Step S200: Obtain image information of the automotive standard parts to be sorted, extract the identification features of the automotive standard parts to be sorted based on the image information, and match the identification features with the standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts.

[0023] The operator places the standard automotive parts to be sorted in the feeding area. An AI camera is positioned directly above the feeding port; preferably, the AI ​​camera is 200mm above the feeding platform. The AI ​​camera captures image information of the standard automotive parts to be sorted and sends the captured image information to the image recognition module. The image recognition module processes the image information and extracts the identification features of the standard automotive parts to be sorted. These identification features include one or more of the following: appearance contour, dimensional parameters, thread specifications, structural features, and surface condition features.

[0024] ; ; In the formula, S represents the comprehensive identification matching degree of the standard part, and n represents the total dimension of the identification features. Let be the weight coefficient of the i-th identification feature. Let be the matching coefficient of the i-th identification feature.

[0025] Step S300: Control the opening and closing state of the access gate set on the feeding path according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches the preset matching threshold, control the access gate to open, so that the standard parts of the automobile to be sorted can enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, control the access gate to close.

[0026] Specifically, the image recognition module compares the calculated overall recognition matching degree of the standard parts with a preset matching threshold. The preset matching threshold is preferably 98%. When the overall recognition matching degree of the standard parts reaches the preset matching threshold, the access gate set on the feeding path is opened, allowing the standard automotive parts to be sorted to enter the sorting channel. When the overall recognition matching degree of the standard parts does not reach the preset matching threshold, the access gate remains closed, thereby preventing non-target standard parts, irregularly shaped parts, damaged parts, or standard parts that do not meet specifications from entering the subsequent sorting system.

[0027] When the access gate closes, the alarm unit simultaneously outputs an audible and visual alarm signal, such as a buzzer and a red indicator light, prompting operators to replace the parts with the correct specifications or qualified quality automotive standard parts before re-feeding. This step allows for the identification and screening of automotive standard parts at the feeding input, preventing incorrect standard parts from entering the subsequent counting and sorting process.

[0028] In step S400, the standard automotive parts passing through the entry gate are conveyed to the magnetic rotary sorting mechanism, which attracts and carries the standard automotive parts. Specifically, after the standard automotive parts to be sorted pass through the entry gate, they enter the working area of ​​the magnetic rotary sorting mechanism. The drive motor, under the control of the motor drive module, drives the magnetic rotary disc to rotate at a constant speed. The rotation speed of the magnetic rotary disc can be adjusted according to the specifications of the standard parts and the production cycle, preferably within a range of 10 r / min to 30 r / min.

[0029] The permanent magnets or magnetic components on the back of the magnetic rotating disc generate a magnetic attraction, stably adhering the automotive standard parts entering the sorting area to the surface or near the surface of the magnetic rotating disc, and rotating synchronously with it. This magnetic adsorption conveying method reduces the occurrence of falling, shifting, stacking, and disorderly flipping of automotive standard parts during transport, improving the stability and continuity of standard part conveying.

[0030] In step S500, when the standard automotive parts move to the guide limit position with the magnetic rotary sorting mechanism, the number of standard automotive parts passing through is limited by the guide limit structure, so that the standard automotive parts enter the falling guide channel in a single-piece state.

[0031] Specifically, in this embodiment, the guide and limiting structure is disposed on the conveying path of the magnetic rotating disk, and may include one or more of the following: a stop, a limiting edge, a guide groove, a single-piece passage, or a height and width limiting structure. The guide and limiting structure constrains the posture, position, and number of automotive standard parts passing through, ensuring that only one automotive standard part is allowed to pass through the guide and limiting position at any given time.

[0032] When two or more automotive standard parts are close to the guide limit position, the guide limit structure can block the stacked or side-by-side automotive standard parts, preventing them from entering the falling guide channel simultaneously, thus creating a physical one-way flow restriction effect. This structure can avoid miscounting problems caused by stacked materials, continuous materials, or multiple parts falling at the same time.

[0033] After a single automotive standard part passes the guide limiting structure, the magnetic rotating disk continues to rotate, moving the part to the entrance of the falling guide channel. At this position, the magnetic attraction on the part weakens, or the part detaches from the magnetic rotating disk under its own weight and the combined action of the guide structure, entering the falling guide channel. The falling guide channel is aligned with the tangential direction of the top of the magnetic rotating disk and is preferably inclined at 30° relative to the horizontal plane to ensure the smooth falling of the automotive standard part and its passage through the detection area.

[0034] In step S600, a detection sensor located at the drop guide channel detects the passage signal of a standard automotive part and generates a real-time cumulative sorting quantity based on the passage signal. Preferably, the detection sensor is located at the drop guide channel and is preferably an infrared sensor. The infrared sensor is embedded in the inner wall of the drop guide channel and forms a 15° angle with the axis of the drop guide channel to stably trigger the detection signal when a standard automotive part passes through the detection area. When a single standard automotive part passes through the drop guide channel, the infrared sensor generates a passage signal and sends the passage signal to the counting controller. The counting controller updates the real-time cumulative sorting quantity based on the received passage signal.

[0035] During the counting process, the counting controller can store the preset sorting quantity, real-time cumulative sorting quantity, identification results, alarm information and completion time for each sorting cycle to form a sorting counting record, which facilitates quality traceability during the automobile assembly process.

[0036] Step S700: Compare the real-time cumulative sorting quantity with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, control the magnetic rotary sorting mechanism to stop running, so as to complete the quantitative counting and sorting of automotive standard parts in one operation.

[0037] In this step, the counting controller dynamically compares the real-time cumulative sorting quantity with the preset sorting quantity. When the real-time cumulative sorting quantity is less than the preset sorting quantity, the counting controller controls the drive motor to keep running, allowing the magnetic rotating disk to continue conveying automotive standard parts; when the preset sorting quantity is reached, the counting controller sends a stop signal to the motor drive module, which then controls the drive motor to stop running, causing the magnetic rotating disk to stop rotating, thus completing one quantitative counting and sorting of automotive standard parts.

[0038] After completing one round of quantitative counting and sorting, the automotive standard parts fall into the receiving tray. When the operator presses down on the receiving tray and removes the quantitatively sorted automotive standard parts, the normally closed limit switch at the bottom of the receiving tray is triggered and generates a restart signal. The restart signal is sent to the counting controller or motor drive module, causing the drive motor to restart, the magnetic rotating disc to resume rotation, and the device enters the next round of automotive standard parts counting and sorting cycle. Thus, the device can realize a closed-loop operation of identification and screening, automatic feeding, single-piece limit, infrared counting, quantitative shutdown, and part retrieval restart.

[0039] In terms of abnormal handling, when the image recognition module determines that the comprehensive recognition matching degree of the standard part does not reach the preset matching threshold, the access gate remains closed, the alarm unit outputs a buzzer alarm and a red indicator signal, and the operator replaces the correct automotive standard part and refeeds the material; when the power supply unit, drive motor, detection sensor or counting controller malfunctions, the control unit triggers the corresponding protection mechanism, stops the drive motor and outputs a fault prompt, and the operation can be resumed after the fault is cleared and the device is restarted.

[0040] In summary, the AI-based automatic identification method for counting and sorting automotive standard parts in the above embodiments of the present invention combines AI image recognition, electromagnetic access control, magnetic rotary sorting, guided and limited single-piece conveying, and infrared sensing counting to achieve closed-loop automated operation of automotive standard parts from identification and screening, automatic feeding, single-piece sorting, accurate counting to quantitative shutdown. Compared with manual sorting and counting, the present invention can effectively improve the sorting and counting efficiency of automotive standard parts, reduce the problems of mis-dispensing, omission, under-counting, and over-counting caused by manual identification and counting, and intercept non-target standard parts or abnormal standard parts from the source through identification matching degree judgment and abnormal alarm, thereby reducing the risk of mis-assembly and omission during the automotive assembly process. At the same time, by dynamically comparing the preset sorting quantity with the real-time cumulative quantity, flexible quantitative supply of standard parts required for different assembly processes can be achieved. With the AC / DC multi-source power supply structure, the adaptability and ease of use of the device in automotive assembly workshops, repair stations, and temporary operation scenarios are improved, thereby improving the accuracy, continuity, and traceability of automotive standard parts sorting operations.

[0041] Example 2 like Figure 2 As shown, the second embodiment of the present invention provides an AI-based automatic identification system for counting and sorting standard automotive parts, comprising: Data module 100 is used to establish a feature database for automotive standard parts, which stores the appearance features, size features and thread specifications of automotive standard parts to be sorted. The feature matching module 200 is used to acquire image information of the automotive standard parts to be sorted, extract the identification features of the automotive standard parts to be sorted based on the image information, and match the identification features with the standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts. The access module 300 is used to control the opening and closing state of the access gate set on the feeding path according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches a preset matching threshold, the access gate is controlled to open, allowing the automotive standard parts to be sorted to enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, the access gate is controlled to close. The conveying module 400 is used to convey the automotive standard parts that pass through the access gate to the magnetic rotary sorting mechanism, and to use the magnetic rotary sorting mechanism to attract and follow the automotive standard parts. The limiting module 500 is used to limit the number of automotive standard parts passing through the guide limiting structure when the automotive standard parts move to the guide limiting position with the magnetic rotary sorting mechanism, so that the automotive standard parts enter the falling guide channel in a single piece state. The counting module 600 is used to detect the passing signal of the automotive standard parts by means of a detection sensor set at the falling guide channel, and generate a real-time cumulative sorting quantity based on the passing signal; The quantitative module 700 is used to compare the real-time cumulative sorting quantity with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, the magnetic rotary sorting mechanism is controlled to stop running to complete the quantitative counting and sorting of automotive standard parts.

[0042] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for counting and sorting standard automotive parts based on AI-based automatic identification, characterized in that, Includes the following steps: Establish an automotive standard parts feature database, which stores the appearance features, size features, and thread specifications of automotive standard parts to be sorted. Image information of automotive standard parts to be sorted is obtained, and the identification features of automotive standard parts to be sorted are extracted based on the image information. The identification features are then matched with standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts. The opening and closing state of the access gate set on the feeding path is controlled according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches the preset matching threshold, the access gate is controlled to open, so that the standard parts of the automobile to be sorted can enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, the access gate is controlled to close. The standard automotive parts passing through the entry gate are transported to the magnetic rotary sorting mechanism, which uses the magnetic rotary sorting mechanism to attract and transport the standard automotive parts. When the standard automotive parts move to the guide limit position with the magnetic rotary sorting mechanism, the number of standard automotive parts passing through is limited by the guide limit structure, so that the standard automotive parts enter the falling guide channel in a single piece state. The detection sensor installed at the falling guide channel detects the passing signal of the automotive standard parts, and generates a real-time cumulative sorting quantity based on the passing signal; The real-time cumulative sorting quantity is compared with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, the magnetic rotary sorting mechanism is controlled to stop running to complete a quantitative counting and sorting of automotive standard parts.

2. The method for counting and sorting automotive standard parts based on AI automatic identification according to claim 1, characterized in that, The formula for calculating the overall matching degree of the standard parts is as follows: ； ； In the formula, S represents the comprehensive identification matching degree of the standard part, and n represents the total dimension of the identification features. Let be the weight coefficient of the i-th identification feature. Let be the matching coefficient of the i-th identification feature.

3. The method for counting and sorting automotive standard parts based on AI automatic recognition according to claim 1, characterized in that, The magnetic rotary sorting mechanism includes a magnetic rotary disk driven by a drive motor. The back of the magnetic rotating disk is provided with a magnetic structure, so that the automotive standard parts entering the magnetic rotating sorting mechanism are attracted to the magnetic rotating disk and rotate with the magnetic rotating disk to the guide limiting structure.

4. The method for counting and sorting automotive standard parts based on AI automatic recognition according to claim 3, characterized in that, When the standard automotive component rotates to the guide limiting structure, the guide limiting structure constrains the posture and passing position of the standard automotive component, so that only one standard automotive component is allowed to pass at any given time. When the standard automotive component continues to rotate to the entrance position of the falling guide channel, the standard automotive component is released from the magnetic attraction of the magnetic structure and enters the falling guide channel under the action of gravity.

5. The method for counting and sorting automotive standard parts based on AI automatic identification according to claim 1, characterized in that, The detection sensor is an infrared sensor; When the standard automotive component passes through the falling guide channel, the infrared sensor is triggered to generate a passing signal; The counting controller updates the real-time cumulative sorting quantity according to the signal, and controls the drive motor to keep running when the real-time cumulative sorting quantity is less than the preset sorting quantity, and controls the drive motor to stop running when the real-time cumulative sorting quantity is equal to the preset sorting quantity.

6. The method for counting and sorting automotive standard parts based on AI automatic identification according to claim 1, characterized in that, The preset sorting quantity is set by a counting controller, and the preset sorting quantity is a positive integer between 1 and 100; Once a quantitative counting and sorting cycle is completed, in response to the part-picking signal generated by the pressing down of the receiving tray, the magnetic rotary sorting mechanism is restarted to enter the next cycle of automotive standard parts counting and sorting.

7. The method for counting and sorting automotive standard parts based on AI automatic recognition according to claim 1, characterized in that, The method further includes: Select the power supply mode, which includes AC power supply mode and DC power supply mode; The input voltage is converted to the device's rated operating voltage via a power conversion module and a voltage regulation module. After power is turned on, the image acquisition unit, drive motor and counting unit perform self-tests; When all inspection units are in normal condition, a workable prompt signal is output; When any self-test unit is in an abnormal state, it outputs a fault warning signal and restricts the start of sorting operations.

8. A system for implementing the AI-based automatic identification method for counting and sorting automotive standard parts as described in any one of claims 1-7, characterized in that, include: The data module is used to establish a database of automotive standard parts features, which stores the appearance features, size features, and thread specifications of automotive standard parts to be sorted. The feature matching module is used to acquire image information of automotive standard parts to be sorted, extract the identification features of automotive standard parts to be sorted based on the image information, and match the identification features with the standard features in the automotive standard parts feature database to obtain the comprehensive identification matching degree of the standard parts. The access module is used to control the opening and closing state of the access gate set on the feeding path according to the comprehensive identification and matching degree of the standard parts. When the comprehensive identification and matching degree of the standard parts reaches the preset matching threshold, the access gate is controlled to open, allowing the automotive standard parts to be sorted to enter the sorting channel. When the comprehensive identification and matching degree of the standard parts does not reach the preset matching threshold, the access gate is controlled to close. The conveying module is used to convey the standard automotive parts that pass through the entry gate to the magnetic rotary sorting mechanism, which uses the magnetic rotary sorting mechanism to attract and follow the standard automotive parts. The limiting module is used to limit the number of automotive standard parts passing through when the automotive standard parts move to the guide limiting position with the magnetic rotary sorting mechanism, so that the automotive standard parts enter the falling guide channel in a single piece state. The counting module is used to detect the passing signal of the automotive standard parts through the detection sensor set at the falling guide channel, and generate a real-time cumulative sorting quantity based on the passing signal; The quantitative module is used to compare the real-time cumulative sorting quantity with the preset sorting quantity. When the real-time cumulative sorting quantity reaches the preset sorting quantity, the magnetic rotary sorting mechanism is controlled to stop running to complete the quantitative counting and sorting of automotive standard parts.

Images