A power board card warehousing detection method and system supporting back-end automatic sorting

Abstract

The application discloses a kind of to support back-end automatic sorting of power board card warehousing detection method related to the technical field of power equipment intelligent warehousing and visual detection.The application is realized by material box scanning code identification, type area division, two-dimensional code reading and whitelist filtering, multi-level brightness Blob analysis, slot and width determination, cross-slot continuation determination, total number verification and data binding upload, and realizes the full-process automation detection of power board card warehousing.The application is compatible with different types of material box, supports different slot and width detection, can still accurately determine whether the board card is occupied or not when barcode identification fails, determines the position information of the target board card to be sorted, and sorts the target board card according to the position information.The application solves the pain points of traditional warehousing detection, such as relying on manual, easy to miss detection, unable to adapt to multi-specification board card, and unable to link automatic sorting, significantly improves the efficiency, accuracy and intelligent level of power board card warehousing.

Description

Technical Field

[0001] This invention relates to a method and system for detecting incoming power circuit boards that supports automated back-end sorting, belonging to the field of intelligent warehousing and visual inspection technology for power equipment. Background Technology

[0002] In the fields of power automation, relay protection, and smart substations, power boards (CPU boards, AC boards, input / output boards, power supply boards, etc.) are diverse in type and specifications, requiring extremely high accuracy in traceability and sorting. Traditional warehousing methods rely on manual barcode scanning, manual counting, and manual location recording, which presents the following prominent problems:

[0003] 1. Low manual efficiency, prone to fatigue, frequent occurrences of misclassification, omission, duplicate counting, and misjudgment of cross-slot boards;

[0004] 2. Incompatible with 6U / 4U hoppers and single / double / triple slot boards, making cable replacement and adjustment complex;

[0005] 3. When barcode reading fails due to damage, obstruction, or angle deviation, it is impossible to determine whether the board is present or occupied.

[0006] 4. Inbound data cannot be linked with the automated sorting mechanism, and manual intervention is still required for shelving and picking;

[0007] 5. The lack of a whitelist filtering mechanism makes it easy to misidentify sub-boards, accessories, and non-target models as boards to be put into the warehouse;

[0008] 6. Unable to output slot number, number of occupied slots, and width information, thus failing to support accurate automatic capture.

[0009] Existing visual inspection solutions mostly only achieve presence or absence detection or simple code reading, without forming a complete closed loop of material bin recognition, typed ROI, code reading and whitelisting, blob size determination, cross-slot continuation judgment, total count verification, data binding, and sorting drive. This makes it difficult to meet the high reliability, high compatibility, and fully automated warehousing and sorting requirements of power boards. Summary of the Invention

[0010] The purpose of this invention is to propose a method and system for detecting power circuit boards entering the warehouse that supports automatic back-end sorting, aiming to achieve unmanned, high-accuracy, highly compatible, and traceable entry detection.

[0011] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0012] In a first aspect, this invention proposes a method for detecting incoming power circuit boards that supports automated backend sorting, comprising:

[0013] Obtain a pre-generated whitelist of board models;

[0014] Read the bin code, parse the bin code, and determine the bin type;

[0015] The current hopper is photographed based on its type, and the captured image is divided into several regions corresponding to that hopper type.

[0016] Perform code reading detection for each area:

[0017] If the barcode reading is successful, information is extracted from the barcode in the area and the board model is parsed. If the board model exists in the whitelist, the corresponding barcode is filtered out. When the number of filtered barcodes is greater than zero, the first region of interest for single slot width identification is determined, and the slot and width occupied by the board are determined based on the first region of interest. If the number of remaining board models is zero, the detection of the current area ends.

[0018] If the code reading fails, determine the second and third regions of interest used to determine the presence or absence of the board, and determine the presence or absence of the board based on the second and third regions of interest.

[0019] Summarize the total number of boards and the total number of barcodes in all areas: if the two are equal, bind and store the location information of the target board for subsequent automatic sorting; if they are not equal, issue an alarm.

[0020] Further, the step of reading the bin code, parsing the bin code, and determining the bin type includes:

[0021] When the bin reaches the scanning position, the barcode scanner is triggered to read the bin code;

[0022] If the read fails, retry; if it still fails after a predetermined number of retries, issue an alarm.

[0023] If the reading is successful, the bin code is parsed and the bin type is determined. The bin type is either a 6U bin or a 4U bin.

[0024] Further, determining the first region of interest for single slot width identification, and determining the slot and width occupied by the board based on the first region of interest, includes:

[0025] Obtain the location information of the region of interest for barcode recognition;

[0026] Based on the region of interest identified by the barcode, a first region of interest is generated by shifting the barcode horizontally to the left by a preset distance.

[0027] The brightness of the first region of interest is enhanced and the major axis value is obtained based on Blob analysis. The slot and width occupied by the board are determined based on the major axis value.

[0028] Furthermore, the step of determining the slot and width occupied by the board based on the major axis value includes:

[0029] When the major axis value of the first region of interest is less than the first threshold, one slot is occupied, and the board width corresponding to the major axis value is determined.

[0030] When the major axis value of the first region of interest is greater than or equal to the first threshold and less than the second threshold, the first region of interest is expanded to a first preset size, it is determined that the board occupies 2 slots, and the board width corresponding to the major axis value is determined.

[0031] When the major axis value of the first region of interest is greater than or equal to the second threshold, the first region of interest is expanded to a second preset size, it is determined that the board occupies 3 slots, and the board width corresponding to the major axis value is determined.

[0032] The board is an expandable board.

[0033] Further, determining the second and third regions of interest for determining the presence or absence of the board, and determining the presence or absence of the board based on the second and third regions of interest, includes:

[0034] Obtain the location information of the region of interest for barcode recognition;

[0035] Based on the barcode recognition region of interest, a second region of interest is generated in its neighboring area; a third region of interest is generated at a preset distance above the center of the second region of interest, wherein the second region of interest is used to determine whether there is a board at the current position, and the third region of interest is used to determine whether it is a continuation of the previous board across slots;

[0036] Brightness is enhanced in the second and third regions of interest, and the major axis value is obtained based on Blob analysis. The presence or absence of the board is determined based on the major axis value.

[0037] Furthermore, the step of determining the presence or absence of the board based on the major axis value includes:

[0038] When the major axis value of the second region of interest is less than the third threshold, it is determined that there is no board at the current location;

[0039] If the major axis value of the second region of interest is greater than or equal to the third threshold, it is determined that there is a board; if the major axis value of the third region of interest is less than the fourth threshold, it is determined that the board is a new board.

[0040] If the major axis value of the third region of interest is greater than or equal to the fourth threshold, it is determined that the board is a continuation of the previous board and there is no new board at the current position.

[0041] Furthermore, if the values ​​are not equal, an alarm will be triggered, including:

[0042] If the total number of circuit boards does not match the total number of barcodes, at least one retake will be performed. If the retake still does not match, an alarm will be triggered.

[0043] Furthermore, the step of binding and storing the location information of the target board for subsequent automatic sorting includes:

[0044] The barcodes, occupied slots, and slot widths of each board are bound according to their serial numbers to generate summary information. The bin codes and the summary information are then uploaded to the upper-level control system. Orders are matched based on the summary information, and the sorting mechanism is driven to the corresponding position for secondary photo positioning and sorting.

[0045] Secondly, this invention proposes a power board warehousing detection system that supports automatic back-end sorting, comprising:

[0046] The database module is used to obtain a pre-generated whitelist of board models;

[0047] The barcode scanning trigger module is used to read the bin code, parse the bin code, and determine the bin type;

[0048] The vision acquisition module is used to trigger the taking of a picture of the current bin based on the bin type;

[0049] The region division module is used to divide the captured image into several regions corresponding to the material bin type according to the material bin type.

[0050] The QR code reading module is used to perform code reading and detection for each area;

[0051] The slot and width determination module is used to extract information from the barcodes in the area where the barcode has been successfully read and parse the board model. If the board model exists in the whitelist, the corresponding barcode is filtered out. When the number of filtered barcodes is greater than zero, the first region of interest for single slot width recognition is determined, and the slot and width occupied by the board are determined based on the first region of interest. If the number of remaining board models is zero, the detection of the current area ends.

[0052] The cross-slot judgment module is used to determine the second and third regions of interest when code reading fails, and to determine whether the board exists based on the second and third regions of interest.

[0053] The data verification module is used to summarize the total number of boards and barcodes in all areas, and to perform total comparison, re-photo verification, and anomaly alarm.

[0054] The communication interaction module is used to determine the location information of the target board to be sorted, and to sort the target board according to the location information.

[0055] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0056] 1. This invention includes automatic bin identification, specification-based visual inspection, barcode parsing and whitelist filtering, automatic determination of board presence / slot / width, cross-slot deduplication, and total quantity consistency verification. It also binds and uploads bin codes and board information, directly driving the automatic sorting mechanism to operate precisely. This achieves full automation from scanning, detection, judgment, verification to uploading and sorting, without human intervention, significantly reducing labor costs and the risk of human error, and achieving the goal of unmanned, high-accuracy, highly compatible, and traceable inbound inspection.

[0057] 2. It supports 6U / 4U bins, single / double / triple slot boards, and multiple width specifications. No mechanical adjustment is required when changing models, and it has strong compatibility.

[0058] 3. Even if barcode recognition fails, it can still accurately determine whether the board has a cross-slot continuity, avoiding missed detections and misjudgments, and has high robustness.

[0059] 4. This invention uses a whitelist filter to automatically remove sub-boards, accessories, and non-target models, thereby improving the purity and accuracy of incoming data.

[0060] 5. The system of this invention outputs bin code, serial number, barcode, slot position and width information, which directly links with automatic sorting equipment to realize closed-loop management of detection and sorting, forming a complete data traceability and business-driven chain.

[0061] 6. This invention uses Blob analysis to stably determine the long axis direction, a reliable cross-slot deduplication mechanism, and error prevention through verification of the total number of incoming cards, which significantly improves the accuracy and efficiency of power board incoming inspection and meets stringent incoming quality control requirements. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of ROI region division for power board barcode recognition and ROI region division for width recognition;

[0063] Figure 2 This is a schematic diagram showing the division of ROI areas for power circuit boards with and without physical object recognition.

[0064] Figure 3 This is a flowchart of the overall workflow for power circuit board receiving and inspection.

[0065] Figure 4 This is a logic diagram for determining the slot and board width after a successful code reading of a power board.

[0066] Figure 5 This is a flowchart of the board identification and judgment process under the condition of power board code reading failure;

[0067] Figure 6This is a diagram showing the power board stored in a 4U container;

[0068] Figure 7 This is a diagram showing the power board stored in a 6U container;

[0069] Attached labels: 1. 4U type bin; 2. 6U type bin; 3. 4U board; 4. 6U board; 5. Barcode recognition ROI; 6. Single slot width recognition ROI; 7. Double slot width recognition ROI; 8. Triple slot width recognition ROI; 9. Board presence / absence judgment ROIⅠ; 10. Board presence / absence judgment ROIⅡ. Detailed Implementation

[0070] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use.

[0071] Example 1:

[0072] This embodiment proposes a method for detecting incoming power circuit boards that supports automated backend sorting. Its workflow diagram is shown below. Figure 3 As shown, it includes:

[0073] Read the board and sub-board models from the database, and the whitelist of board models to be filtered;

[0074] When the hopper reaches the scanning position, the barcode scanner is triggered to read the hopper code; if the reading fails, it is retried; if it still fails after two retries, an alarm is triggered; if the reading is successful, the hopper code is parsed to determine whether the hopper type is 6U hopper 2 or 4U hopper 1. Figure 1 As shown, the 4U-shaped storage bin 1 can store 30 power 4U boards 3, and its storage diagram is as follows. Figure 6 As shown, it can recognize up to 30 single-slot boards; the 6U-shaped hopper 2 can store 20 power 6U boards 4, and its storage diagram is shown below. Figure 7 As shown, it can recognize up to 20 single-slot cards;

[0075] The camera is triggered to take a picture based on the bin type and the corresponding recognition program is run. The picture image is divided into barcode recognition ROI 5 according to the bin type. Specifically, the 6U bin 2 is divided into 20 barcode recognition ROI 5, numbered in the order of 1-20, and corresponds one-to-one with the serial number of the single slot board in the 4U bin 1; the 4U bin 1 is divided into 30 barcode recognition ROI 5, numbered in the order of 1-30, and corresponds one-to-one with the serial number of the single slot board in the 6U bin 2.

[0076] The system enters different detection branches according to the partition and performs partition detection. Each barcode recognition ROI 5 performs a QR code reading operation.

[0077] If the code reading is successful, such as Figure 4 As shown, extract all barcode information within the area and parse the board model. If the board model exists in the whitelist, filter out the corresponding barcode.

[0078] When the number of filtered barcodes is greater than zero, set the number of barcodes in the current area to the number of filtered results, and set the number of boards to 1; using the barcode position in this area as a reference, establish a single-slot width recognition ROI of 100px × 360px 10 pixels to the left, such as... Figure 1 As shown, brightness enhancement is applied to this area to expand the board outline and improve the stability of Blob extraction. Blob analysis is performed within the brightness range of 240-255 to obtain the major axis value of the bright area of ​​the board, and the occupied slot and board width are determined based on the major axis value;

[0079] When the major axis value is less than 270mm, it occupies one slot, and the board width is generally 200mm / 250mm. When the major axis value is in the range of 180mm~225mm, the corresponding board width is 200mm, and when the major axis value is in the range of 225mm~270mm, the corresponding board width is 250mm.

[0080] When the major axis is ≥270mm, the recognition area is expanded to a dual-slot width recognition ROI 7, occupying 2 slots. The board width is generally 300mm / 400mm / 500mm. When the major axis value is in the range of 280mm~320mm, the corresponding board width is 300mm; when the major axis value is in the range of 380mm~420mm, the corresponding board width is 400mm; and when the major axis value is in the range of 480mm~520mm, the corresponding board width is 500mm.

[0081] When the long axis is ≥520mm, the recognition area will be expanded to a three-slot width recognition ROI 8, occupying 3 slots. When the long axis value is in the range of 730mm~770mm, the corresponding board width is 750mm, and it supports expansion to other width specifications.

[0082] When the number of filtered barcodes is zero, both the number of barcodes and the number of cards in the current area are set to 0, and the recognition of the current area ends.

[0083] If reading the code fails, such as Figure 5 As shown, based on the barcode position in this area, a 150px × 360px ROI (Region of Interest) for card presence / absence determination is created. Then, a 150px × 220px ROI (Region of Interest) for card presence / absence determination is created centered 160px above the center of ROI (Region of Interest) ... Figure 2As shown, ROIⅠ 9, which determines whether a board exists in the current area, and ROIⅡ 10, which determines whether the board in the current area is a continuation of the previous board across slots, to avoid duplicate counting.

[0084] Brightness enhancement was applied to both regions to expand the board's outline and improve the stability of blob extraction. Blob analysis was performed within the brightness range of 240–255 to obtain the major axis value of the highlighted areas of the board.

[0085] If the board has no value for the major axis of ROIⅠ 9 or the major axis value is <190mm, then it is determined that there is no board and the board quantity is set to 0.

[0086] If the board does not determine whether the major axis value of ROIⅠ 9 is ≥190mm, it is determined that there is a board in that area. Then, it is further determined whether the board determines whether the major axis value of ROIⅡ 10 is greater than 200mm.

[0087] If the major axis value is less than 200mm, the board in this area is determined to be a new board, and the number of boards in the current area is set to 1.

[0088] If the major axis value is ≥200mm, the board in this area is determined to be a continuation of the previous board, and the number of boards in the current area is set to 0.

[0089] Summarize the total number of circuit boards and barcodes across all regions:

[0090] If the total number of boards is equal to the total number of barcodes, the barcode, slot, and width are bound by serial number as summary information. The bin code and summary information are uploaded to the host computer. After receiving the bin code and board information, the host computer matches the slot serial number according to the barcode or board model, sends the slot serial number to the automatic sorting mechanism, drives the sorting mechanism to the corresponding position for secondary photo positioning and performs grabbing.

[0091] If the total number of circuit boards does not match the total number of barcodes, a new photo verification will be performed. If they still do not match, an alarm will be triggered and manual intervention will be requested.

[0092] In this embodiment, the purpose of brightness enhancement is to improve the stability of Blob analysis in extracting high-brightness areas, while increasing the range of the major axis judgment value to cope with the object image expansion caused by brightness enhancement and improve the accuracy of judgment. All Blob analyses use a fixed threshold of 240~255 to extract high-brightness areas to ensure the consistency between the board outline and the major axis measurement.

[0093] Example 2:

[0094] This embodiment proposes a code reading scenario based on the first embodiment, where a 4U specification hopper occupies one slot.

[0095] After the system starts, it reads all board and sub-board models from the database and generates a whitelist library to filter sub-boards, accessory boards, and non-target models.

[0096] The material bin arrives at the inspection station via the conveyor line. The sensor triggers the barcode scanner to read the material bin code. If the reading is successful, the material bin is analyzed. The material bin type is 4U type 1, corresponding to 30 slots, which can store 30 power 4U boards 3.

[0097] The system triggers the camera to take a picture and obtain a top view image of the bin. According to the 4U bin rule, the image is automatically divided into 30 barcode recognition ROIs and numbered sequentially from 1 to 30.

[0098] Perform QR code recognition on each barcode recognition ROI 5 in sequence. When the 12th barcode recognition ROI 5 is executed, the QR code is successfully read and the barcode string is obtained.

[0099] The barcode is parsed to extract the board model, which is then compared with the whitelist. If the board is not in the whitelist, it is determined to be a valid board to be put into storage, and the quantity of barcodes in area 12 and the quantity of boards are both set to 1.

[0100] Using the barcode position in region 12 as a reference, offset 10 pixels to the left to establish a single-slot width ROI 6 with a size of 100px × 360px. Perform brightness enhancement on this region to expand the board outline and improve the stability of Blob extraction. Perform Blob analysis within the brightness range of 240 to 255 to obtain the major axis value of the board's bright area. The major axis value is 245mm, which satisfies: 180mm ≤ 245mm < 270mm. It is determined that the board occupies 1 slot with a width of 250mm.

[0101] Continue to complete the detection of the remaining barcode recognition ROI 5. After all areas have been detected, summarize the total number of boards and the total number of barcodes. If the numbers are consistent, the verification is successful. Bind the bin code with the information of area 12 (barcode, slot 1, 250mm) and upload it to the host computer.

[0102] The host computer matches the order based on the barcode and issues the instruction for slot number 12 to the automatic sorting mechanism.

[0103] The sorting mechanism moves above slot number 12, performs secondary precision positioning, and completes the board grabbing.

[0104] Example 3:

[0105] This embodiment proposes a code reading scenario based on the first embodiment, where a 6U specification hopper occupies two slots.

[0106] Get the pre-generated whitelist;

[0107] The material bin arrives at the inspection station via the conveyor line. The sensor triggers the barcode scanner to read the material bin code. If the reading is successful, the material bin is analyzed. The material bin type is 6U type 2, which corresponds to 20 slots and can store 20 power 6U boards 4.

[0108] The system triggers the camera to take a picture and obtain a top view image of the bin. According to the 6U bin rule 2, the image is automatically divided into 20 barcode recognition ROIs 5 and numbered sequentially from 1 to 20.

[0109] Perform QR code recognition on each barcode recognition ROI 5 in sequence. When the 18th barcode recognition ROI 5 is executed, the QR code is successfully read and the barcode string is obtained.

[0110] The barcode is parsed to extract the board model, which is then compared with the whitelist. If the board is not in the whitelist, it is determined to be a valid board to be put into storage, and the quantity of barcodes in area 18 and the quantity of boards are both set to 1.

[0111] Using the barcode position in region 12 as a reference, offset 10 pixels to the left to create a single-slot width recognition ROI 6 with a size of 100px × 360px. Perform brightness enhancement on this region and perform Blob analysis within the brightness range of 240 to 255 to obtain the major axis value of the high-brightness area of ​​the board. If the major axis value is 405mm > 270mm, expand the recognition area to a double-slot width recognition ROI 7, determine that the board occupies 2 slots, 380mm ≤ 405mm < 420mm, and the width is 400mm.

[0112] Continue to complete the inspection of the remaining areas. After all areas have been inspected, summarize the total number of boards and the total number of barcodes. If the numbers are consistent, the verification is successful. Bind the bin code to the information of area 18 (barcode, slot 2, 405mm) and upload it to the host computer.

[0113] The host computer issues a sorting instruction, and the sorting mechanism performs an open-type gripping action according to the width of the double slots.

[0114] Example 4:

[0115] This embodiment proposes a code reading scenario based on the first embodiment, where a 4U specification hopper occupies 3 slots.

[0116] Get the whitelist;

[0117] The hopper arrives at the inspection station via the conveyor line. The reading is successful. The hopper is analyzed and identified as a 4U type hopper 1, corresponding to 30 slots.

[0118] The system triggers the camera to take a picture and obtain a top view image of the bin. According to the 4U bin rule, the image is automatically divided into 30 barcode recognition ROIs and numbered sequentially from 1 to 30.

[0119] Perform QR code recognition on each barcode recognition ROI 5 in sequence. When the QR code is successfully read when the 7th barcode recognition ROI 5 is executed, the barcode string is obtained.

[0120] The barcode is parsed to extract the board model, which is then compared with the whitelist. If the board is not in the whitelist, it is determined to be a valid board to be put into storage, and the quantity of barcodes in area 7 is set to 1, and the quantity of boards is also set to 1.

[0121] Using the barcode position in region 7 as a reference, offset 10 pixels to the left to create a single-slot width recognition ROI 6 with a size of 100px × 360px. Perform brightness enhancement on this region and perform Blob analysis within the brightness range of 240 to 255 to obtain the major axis value of the high-brightness area of ​​the board. If the major axis value is 755mm > 520mm, expand the recognition area to a three-slot width recognition ROI 8. Determine that the board occupies 3 slots, 730mm ≤ 755mm < 770mm, with a width of 750mm.

[0122] Continue to complete the detection of the remaining 29 barcode recognition ROI 5. After all areas have been detected, summarize the total number of boards and the total number of barcodes. If the numbers are consistent, the verification is successful. Bind the bin code with the information of area 7 (barcode, slot 3, 750mm) and upload it to the host computer.

[0123] The host computer matches the order based on the barcode and sends the instruction for slot number 7 to the automatic sorting mechanism. The sorting mechanism performs a large-stroke gripping operation according to the width of the three slots.

[0124] Example 5:

[0125] This embodiment proposes a scenario where reading the code of any material box fails, and an independent circuit board exists, based on embodiment one.

[0126] The system sequentially performs QR code recognition for each barcode recognition ROI 5. When it reaches the 9th barcode recognition ROI 5, the QR code reading fails, the barcode count is set to 0, and the system automatically enters the barcode reading failure branch process.

[0127] Based on the barcode position in this area, create a board with a size of 150px×360px for determining whether the board has a judgment ROIⅠ9, and create a board with a size of 150px×220px for determining whether the board has a judgment ROIⅡ 10 with a size of 150px×220px, offset 160px upward from the center of the board with a judgment ROIⅠ 9.

[0128] Brightness enhancement and Blob analysis were performed on ROIⅠ 9 and ROIⅡ 10, which are used to determine whether a board is present or not. The major axis value of ROIⅠ 9 was measured to be 225mm, which is greater than 190mm, indicating that a board is present in the current area.

[0129] Continue to determine if the board has a ROIⅡ 10 long axis value of 175mm, which is less than 200mm. This board is an independent board and is not a continuation of the previous board across the slot. Set the number of boards in area 9 to 1.

[0130] After proceeding to the next area for testing, the final total count is verified to be consistent, and the data entry is completed.

[0131] Example 6:

[0132] This embodiment, based on embodiment one, proposes a case where reading the code of any material box fails and there is a case where the card is continued across slots.

[0133] The system sequentially performs QR code recognition for each barcode recognition ROI 5. When it reaches barcode recognition ROI 15, the QR code reading fails, the barcode count is set to 0, and the system automatically enters the barcode reading failure branch process.

[0134] Based on the barcode position in this area, create a board with a size of 150px×360px for determining whether the board has a judgment ROIⅠ9, and create a board with a size of 150px×220px for determining whether the board has a judgment ROIⅡ 10 with a size of 150px×220px, offset 160px upward from the center of the board with a judgment ROIⅠ 9.

[0135] Brightness enhancement and Blob analysis were performed on ROIⅠ 9 and ROIⅡ 10, respectively, to determine whether a board exists. The major axis value of ROIⅠ 9 was measured to be 230mm, which is greater than 190mm, indicating that a board exists in the current area.

[0136] Continue to determine if the board has a ROIⅡ 10 major axis value of 220mm, which is greater than 200mm. This board is a continuation of the previous board across the slot, so set the number of boards in area 15 to 1.

[0137] After proceeding to the next area for testing, the final total count is verified to be consistent, and the data entry is completed.

[0138] Example 7:

[0139] This embodiment proposes a total count verification and anomaly alarm based on embodiment one.

[0140] QR code detection is performed on each barcode recognition ROI 5 in sequence. After all detections are completed, the total number of boards and the total number of barcodes in all areas are summarized. 1. The total number of boards = 14, and the total number of barcodes = 13, which are not equal.

[0141] Automatically trigger the camera to take another picture and perform a full process check again;

[0142] If the second test result still shows a discrepancy in quantity, the system will determine it as an abnormal state, trigger an audible and visual alarm, and simultaneously upload the abnormal information to the host computer. The system will then pause the automatic process and wait for manual inspection and confirmation of the board placement.

[0143] After a manual reset, the system restarts the testing process.

[0144] It should be noted that the above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art can make various modifications and improvements without departing from the essence of the technical solution of the present invention, and these modifications and improvements also fall within the scope of protection of the present invention.

Claims

1. A method for detecting incoming power circuit boards that supports automated backend sorting, characterized in that, include: Obtain a pre-generated whitelist of board models; Read the bin code, parse the bin code, and determine the bin type; The current hopper is photographed based on its type, and the captured image is divided into several regions corresponding to that hopper type. Perform code reading detection for each area: If the barcode reading is successful, information is extracted from the barcode in the area and the board model is parsed. If the board model exists in the whitelist, the corresponding barcode is filtered out. When the number of filtered barcodes is greater than zero, the first region of interest for single slot width identification is determined, and the slot and width occupied by the board are determined based on the first region of interest. If the number of remaining board models is zero, the detection in the current area will end; If the code reading fails, determine the second and third regions of interest used to determine the presence or absence of the board, and determine the presence or absence of the board based on the second and third regions of interest. Summarize the total number of boards and the total number of barcodes in all areas: if the two are equal, bind and store the location information of the target board for subsequent automatic sorting; If they are not equal, an alarm will be triggered.

2. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 1, characterized in that, The process of reading the bin code, parsing the bin code, and determining the bin type includes: When the bin reaches the scanning position, the barcode scanner is triggered to read the bin code; If the read fails, retry; if it still fails after a predetermined number of retries, issue an alarm. If the reading is successful, the bin code is parsed and the bin type is determined. The bin type is either a 6U bin or a 4U bin.

3. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 1, characterized in that, The process of determining a first region of interest for single-slot width identification, and determining the slot and width occupied by the board based on the first region of interest, includes: Obtain the location information of the region of interest for barcode recognition; Based on the region of interest identified by the barcode, a first region of interest is generated by shifting the barcode horizontally to the left by a preset distance. The brightness of the first region of interest is enhanced, and the major axis value of the high-brightness region is obtained based on Blob analysis. The slot and width occupied by the board are determined based on the major axis value.

4. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 3, characterized in that, The determination of the slot and width occupied by the board based on the major axis value includes: When the major axis value of the first region of interest is less than the first threshold, one slot is occupied, and the board width corresponding to the major axis value is determined. When the major axis value of the first region of interest is greater than or equal to the first threshold and less than the second threshold, the first region of interest is expanded to a first preset size, it is determined that the board occupies 2 slots, and the board width corresponding to the major axis value is determined. When the major axis value of the first region of interest is greater than or equal to the second threshold, the first region of interest is expanded to a second preset size, it is determined that the board occupies 3 slots, and the board width corresponding to the major axis value is determined. The board is an expandable board.

5. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 1, characterized in that, The step of determining the second and third regions of interest for judging the presence or absence of a board, and determining the presence or absence of a board based on the second and third regions of interest, includes: Obtain the location information of the region of interest for barcode recognition; Based on the barcode recognition region of interest, a second region of interest is generated in its neighboring area; a third region of interest is generated at a preset distance above the center of the second region of interest, wherein the second region of interest is used to determine whether there is a board at the current position, and the third region of interest is used to determine whether it is a continuation of the previous board across slots; The brightness of the second and third regions of interest is enhanced, and based on Blob analysis, the major axis value of the high-brightness region is obtained. The presence or absence of the board is determined based on the major axis value.

6. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 5, characterized in that, The process of determining the presence or absence of a board based on its major axis value includes: When the major axis value of the second region of interest is less than the third threshold, it is determined that there is no board at the current location; If the major axis value of the second region of interest is greater than or equal to the third threshold, it is determined that there is a board; if the major axis value of the third region of interest is less than the fourth threshold, it is determined that the board is a new board. If the major axis value of the third region of interest is greater than or equal to the fourth threshold, it is determined that the board is a continuation of the previous board and there is no new board at the current position.

7. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 1, characterized in that, If the values ​​are not equal, an alarm will be triggered, including: If the total number of circuit boards does not match the total number of barcodes, at least one retake will be performed. If the retake still does not match, an alarm will be triggered.

8. The method for detecting incoming power circuit boards that supports automatic backend sorting as described in claim 1, characterized in that, The step of binding and storing the location information of the target board for subsequent automatic sorting includes: The barcodes, occupied slots, and slot widths of each board are bound according to their serial numbers to generate summary information. The bin codes and the summary information are then uploaded to the upper-level control system. Orders are matched based on the summary information, and the sorting mechanism is driven to the corresponding position for secondary photo positioning and sorting.

9. A power board warehousing inspection system supporting back-end automatic sorting, used to execute the power board warehousing inspection method supporting back-end automatic sorting as described in any one of claims 1-8, characterized in that, include: The database module is used to obtain a pre-generated whitelist of board models; The barcode scanning trigger module is used to read the bin code, parse the bin code, and determine the bin type; The vision acquisition module is used to trigger the taking of a picture of the current bin based on the bin type; The region division module is used to divide the captured image into several regions corresponding to the material bin type according to the material bin type. The QR code reading module is used to perform code reading and detection for each area; The slot and width judgment module is used to extract information from the barcode in the area where the barcode is successfully read and parse the board model. If the board model exists in the whitelist, the corresponding barcode is filtered out. When the number of filtered barcodes is greater than zero, a first region of interest is determined for single slot width recognition, and the slot and width occupied by the board are determined based on the first region of interest. If the number of remaining board models is zero, the detection in the current area will end; The cross-slot judgment module is used to determine the second and third regions of interest when code reading fails, and to determine whether the board exists based on the second and third regions of interest. The data verification module is used to summarize the total number of boards and barcodes in all areas, and to perform total comparison, re-photo verification, and anomaly alarm. The communication interaction module is used to determine the location information of the target board to be sorted, and to sort the target board according to the location information.

Images