X-ray nondestructive testing artificial accurate sorting mistake proofing method based on visual identification

Abstract

The present application relates to product classification technical field, specifically to the X-ray nondestructive testing artificial accurate sorting mistake proofing method based on visual identification, the present application realizes the automatic ordering of detection result by introducing priority sequence and sub-priority sequence, optimizes the sorting process, improves the efficiency of detection and sorting, avoids the sensor wiring and wiring leaving image in the X-ray detection imaging picture, thereby eliminating the interference of detection image reading and discrimination, improving the accuracy of detection, without laying out sensors on each detection tool, reducing the cost of tooling, recording the process video before and after sample taking through industrial camera, and judging the taking order based on priority sequence and process video, reducing the false alarm and missed alarm phenomenon in the manual sorting process, improving the accuracy of sorting, thereby solving the problem of low accuracy of the existing manual sorting mistake proofing method.

Description

Technical Field

[0001] The present invention relates to the technical field of product classification, and in particular to a method for accurate manual sorting and error prevention based on visual recognition and X-ray non-destructive testing. Background Art

[0002] When existing rotary table X-ray nondestructive testing equipment is used to inspect multiple small workpieces in batches, manual sorting based on the results is typically performed after the test is completed. This typically involves using photoelectric or laser sensors on the workpiece to sense the presence of the product, combined with LED lights to indicate the workpiece inspection results. However, in actual use, the following issues arise:

[0003] The sensor needs to be placed on the tooling and brought into the testing room along with the testing platform and tooling. Due to the principle of X-ray non-destructive testing, the sensor's wiring and routing can easily leave an image in the X-ray image, which may interfere with the reading and interpretation of the test image.

[0004] For different samples, the detection tooling is different, and sensors need to be deployed on each tooling.

[0005] Sensors are consumables. After long-term high-frequency X-ray exposure, they are prone to failure and need to be replaced, affecting the equipment's detection and production efficiency.

[0006] During manual unloading and sorting, the movements during the picking process can easily interfere with the sensor sensing, causing false alarms in the sorting reminders. Summary of the Invention

[0007] The purpose of the present invention is to provide a method for accurate manual sorting and error prevention in X-ray non-destructive testing based on visual recognition, aiming to solve the problem of low accuracy of existing manual sorting and error prevention methods.

[0008] To achieve the above objectives, the present invention provides a method for accurate manual sorting and error prevention of X-ray non-destructive testing based on visual recognition, comprising the following steps:

[0009] Detecting samples and outputting priority sequence and sub-priority sequence;

[0010] Recording the process video before and after the sample is taken by an industrial camera;

[0011] Determine the order of picking based on the priority sequence and the process video;

[0012] Cutting materials based on the sub-priority sequence and judging the cutting process;

[0013] After unloading is completed, the status of the material tray is judged to complete the unloading operation.

[0014] Among them, in "testing samples and outputting priority sequence", it includes:

[0015] After testing the sample, the sample is discharged to a designated location;

[0016] A priority sequence and a sub-priority sequence are generated based on the samples.

[0017] The process of “recording a video of the process before and after taking the sample by an industrial camera” includes:

[0018] During the sample taking process, the industrial camera triggers the snapshot function to record the taking process.

[0019] After taking the sample, the industrial camera captures the cutting tooling and records the status of the tooling.

[0020] The process of “determining the order of picking up items based on the priority sequence and the process video” includes:

[0021] Use AI algorithms to identify captured images of blanking tooling to determine whether the material picking and blanking actions are correct;

[0022] If the material picking process is wrong, an alarm will be issued and the sample restoration operation will be output to re-unload the material. If the material picking process is normal, unloading will continue.

[0023] The step of “cutting materials based on the sub-priority sequence and judging the cutting process” includes:

[0024] Collect the current status diagram of the material tray and the view after unloading;

[0025] Based on the AI ​​algorithm and sub-priority sequence, the current status diagram of the material tray and the view after unloading are judged. If it is wrong, an alarm is issued and the sample restoration operation is output to re-unload. If it is correct, unloading continues.

[0026] Among them, "judging the status of the tray after unloading and completing the unloading operation" includes:

[0027] Collect the current status picture of the material tray;

[0028] The current status image is judged and detected. If the tray is empty, the material is unloaded. If the tray is filled with material, an alarm is issued, and the samples in the next priority sequence are restored and the material is unloaded again.

[0029] Among them, in "judging and detecting the current status picture, if there is no material in the material tray, unloading is completed; if there is material in the material tray, an alarm is issued, and the samples in the secondary priority sequence are restored and unloading is started again", the alarm includes the sounding of an alarm buzzer and the flashing of an alarm light.

[0030] The method of the present invention for accurate manual sorting and error prevention of X-ray non-destructive testing based on visual recognition includes the following steps: testing samples and outputting a priority sequence and a sub-priority sequence; recording a process video before and after the sample is taken by an industrial camera; judging the order of taking based on the priority sequence and the process video; unloading materials based on the sub-priority sequence and judging the unloading process; judging the status of the material tray after unloading is completed to complete the unloading operation. By introducing a priority sequence and a sub-priority sequence, the present invention realizes automatic sorting of the test results, optimizes the sorting process, and improves the efficiency of detection and sorting. The photoelectric or laser sensors on traditional tooling are abandoned, and the sensor wiring and routing are prevented from leaving images in the X-ray detection imaging picture, thereby eliminating interference with the detection image reading and discrimination, and improving the accuracy of detection. There is no need to deploy sensors on each inspection tool, which reduces tooling costs and the trouble of re-deploying sensors due to tooling replacement, thereby improving the versatility and flexibility of the equipment. Industrial cameras are used to record process videos before and after sample retrieval, and the order of retrieval is determined based on the priority sequence and process video. This reduces false positives and missed positives in the manual sorting process, improves sorting accuracy, and solves the problem of low accuracy of existing manual sorting error prevention methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0032] Figure 1 This is a flow chart of the method for accurate manual sorting and error prevention in X-ray non-destructive testing based on visual recognition provided by the present invention.

[0033] Figure 2 It is a flow chart for testing samples and outputting priority sequence and sub-priority sequence.

[0034] Figure 3 It is a flow chart of recording the process video before and after the sample is taken by an industrial camera.

[0035] Figure 4 It is a flow chart for determining the picking order based on the priority sequence and the process video.

[0036] Figure 5 It is a flow chart for unloading materials based on the sub-priority sequence and judging the unloading process.

[0037] Figure 6 It is a flow chart for judging the status of the material tray after unloading and completing the unloading operation.

[0038] Figure 7 This is a schematic diagram of the result of the incorrect cutting sequence prompt.

[0039] Figure 8 It is a schematic diagram of the implementation steps of the method for accurate manual sorting and error prevention of X-ray non-destructive testing based on visual recognition provided by the present invention.

[0040] Figure 9 This is a schematic diagram of workpiece detection. DETAILED DESCRIPTION

[0041] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0042] See also Figures 1 to 9 The present invention provides a method for accurate manual sorting and error prevention of X-ray non-destructive testing based on visual recognition, comprising the following steps:

[0043] S1 tests the samples and outputs the priority sequence and sub-priority sequence;

[0044] S11 tests the sample and then discharges the sample to a designated location;

[0045] Specifically, after the sample is tested on the turntable X-ray non-destructive testing equipment, the equipment automatically moves the sample to the preset discharge area based on the test results. The discharge area is equipped with sensors or mechanical devices to ensure that the sample can be accurately and stably placed in the designated position to facilitate subsequent operations.

[0046] S12 generates a priority sequence and a sub-priority sequence based on the samples.

[0047] Based on the test results, the system categorizes samples into a priority sequence and a sub-priority sequence. The priority sequence contains samples that require priority processing, which may have a higher test priority or require more urgent processing. The sub-priority sequence contains the remaining samples, which are processed after the priority sequence is completed.

[0048] S2 records the process video before and after the sample is taken using an industrial camera;

[0049] When S21 is taking samples, the industrial camera triggers the snapshot function to record the taking process.

[0050] Specifically, when the operator begins to pick up a sample, the industrial camera triggers a snapshot function, recording key frames during the process. This snapshot function ensures that key information such as the operator's hand movements and the sample's movement trajectory are captured. This information is used to subsequently determine the correctness of the picking sequence and movement.

[0051] After taking the sample, the S22 industrial camera takes a snapshot of the blanking tooling and records the tooling status.

[0052] Specifically, after a sample is removed, an industrial camera captures a snapshot of the blanking tooling, recording its status. The captured image includes information such as the location and quantity of the sample on the tooling. This information is used to subsequently assess the blanking process.

[0053] S3 determines the picking order based on the priority sequence and the process video;

[0054] S31 uses AI algorithms to identify captured images of blanking tooling to determine whether the material picking and blanking actions are correct;

[0055] Specifically, the AI ​​algorithm performs image recognition and analysis on captured images of the cutting tooling. By comparing them against preset templates or standard images, it determines whether the material handling and cutting actions meet the requirements. If an error or anomaly is detected, the AI ​​algorithm will issue an alarm signal.

[0056] If the material picking process is wrong, S32 will issue an alarm and output the sample restoration operation to re-cut the material. If the material picking process is normal, continue cutting the material.

[0057] Specifically, if the AI ​​algorithm determines that the material handling process is incorrect, the system will issue an alarm through a sounding buzzer and flashing alarm lights. Simultaneously, the system will output sample restoration instructions, guiding the operator to return the sample to its original location or perform other necessary processing. If the material handling process is normal, the system will continue with the unloading operation.

[0058] S4 performs material unloading based on the sub-priority sequence and judges the unloading process;

[0059] S41 collects the current status of the material tray and the view after unloading;

[0060] Specifically, during the unloading process, the industrial camera captures the current state of the tray and the view after unloading in real time. These images are used to judge the unloading process later.

[0061] S42 judges the current status diagram of the material tray and the view after unloading based on the AI ​​algorithm and sub-priority sequence. If it is wrong, it will issue an alarm and output the sample restoration operation to re-unload. If it is correct, it will continue unloading.

[0062] Specifically, the AI ​​algorithm performs image recognition and analysis on the captured images. By comparing them against preset templates or standard images, it determines whether the blanking process meets the requirements. If a blanking error or anomaly is detected, the AI ​​algorithm will issue an alarm signal and output instructions for sample restoration. If the blanking process is normal, the system will continue with subsequent operations.

[0063] After unloading, S5 judges the status of the tray and completes the unloading operation.

[0064] S51 collects the current status picture of the material tray;

[0065] Specifically, after the unloading operation is completed, the industrial camera collects images of the current status of the tray. These images are used for subsequent judgment of the tray status.

[0066] S52 judges and detects the current status image. If the tray is empty, unloading is completed. If the tray is full, an alarm is issued, and the samples in the secondary priority sequence are restored and unloading is restarted.

[0067] The alarm includes the sounding of an alarm buzzer and the flashing of an alarm light.

[0068] Specifically, the AI ​​algorithm performs image recognition and analysis on the collected pictures of the current status of the material tray.

[0069] If the tray is empty, the system assumes the unloading operation is complete and can proceed. If the tray is empty, the system issues an alarm through a sounding buzzer and flashing light. Simultaneously, the system issues instructions for restoring the sample in the lower priority sequence, guiding the operator to return the sample to its original location or perform other necessary processing.

[0070] Beneficial effects:

[0071] 1. No wiring is required on the tooling, and the inspection does not affect the X-ray imaging effect and result reading and judgment.

[0072] 2. High adaptability: It is not limited to the specific shape of the sample workpiece.

[0073] 3. High economic efficiency: Different detection tools are used for different samples, and there is no need to increase the number of sensors on the tools.

[0074] 4. Long service life. Professional industrial cameras are used for visual recognition, which are resistant to various harsh conditions in the equipment production environment. They are non-destructive parts and do not require frequent replacement.

[0075] The above disclosure is only a preferred embodiment of the method for accurate manual sorting and error prevention of X-ray non-destructive testing based on visual recognition of the present invention. Of course, this cannot be used to limit the scope of the rights of the present invention. Ordinary technicians in this field can understand that implementing all or part of the processes of the above embodiment and making equivalent changes in accordance with the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for accurate manual sorting and error prevention for X-ray nondestructive testing based on visual recognition, characterized by: The following steps are involved: Detecting samples and outputting priority sequence and sub-priority sequence; Recording the process video before and after the sample is taken by an industrial camera; Determine the picking order based on the priority sequence and the process video, including: using an AI algorithm to identify the captured image of the blanking tooling to determine whether the picking and blanking actions are correct; if the picking process is wrong, issue an alarm and output a sample restoration operation to re-cut the material; if the picking process is normal, continue cutting the material; The material is unloaded based on the sub-priority sequence, and the unloading process is judged, including: collecting the current state diagram of the material tray and the view after unloading; judging the current state diagram of the material tray and the view after unloading based on the AI ​​algorithm and the sub-priority sequence, if it is wrong, an alarm is issued, and the sample restoration operation is output to re-unload the material; if it is correct, the material is continued; After unloading is completed, the status of the material tray is judged to complete the unloading operation.

2. The method for accurate manual sorting and error prevention for X-ray nondestructive testing based on visual recognition according to claim 1, characterized in that: In "Testing samples and outputting priority sequence", include: After testing the sample, the sample is discharged to a designated location; A priority sequence and a sub-priority sequence are generated based on the samples.

3. The method for accurate manual sorting and error prevention for X-ray nondestructive testing based on visual recognition according to claim 1, characterized in that: "Recording the process video before and after the sample is taken by an industrial camera" includes: During the sample taking process, the industrial camera triggers the snapshot function to record the taking process; After taking the sample, the industrial camera captures the cutting tooling and records the status of the tooling.

4. The method for accurate manual sorting and error prevention for X-ray nondestructive testing based on visual recognition according to claim 1, characterized in that: "Judging the tray status after unloading and completing the unloading operation" includes: Collect the current status picture of the material tray; If the judgment result is that there is no material in the tray, the system considers that the unloading operation has been completed and subsequent operations can be carried out. If the judgment result is that there is material in the tray, the system will issue an alarm by sounding the alarm buzzer and flashing the alarm light. At the same time, the system will output the sample recovery operation instruction for the second priority sequence to guide the operator to put the sample back in place or perform other necessary processing.

Images