Security inspection and sorting integrated system and method based on AI visual inspection

Abstract

The invention belongs to the technical field of automatic control, and discloses a security check and sorting integrated system and method based on AI visual inspection, and the system is characterized in that sorted articles are scanned through a six-surface code scanner, a scanning result is sent to a processing device, an image acquisition device is arranged in a security check instrument, and the processing device is used for processing the scanning result; the processing device receives an image acquired by the image acquisition device and a scanning result of the six-surface code scanner, and issues a sorting instruction to the sorting device; and the sorting device sorts the articles according to the instruction issued by the processing device. According to the invention, the processing device processes the order number information and the image information collected by the six-surface code scanner and the image collection device, automatically and accurately recognizes the type of the contraband, recognizes the confidence and classifies the contraband, can greatly reduce the manpower, reduces the misjudgment, reduces the manpower cost of a transfer center, and improves the safety inspection efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of automatic control technology and relates to an integrated security inspection and sorting system and method based on AI visual detection. Background Art

[0002] With the rapid development of the global economy and the continued prosperity of e-commerce, the express delivery industry has ushered in unprecedented opportunities and challenges. The accompanying express delivery security issues are particularly prominent. The security inspection machines and automated sorting equipment in the existing transfer centers are operated in different locations. The security inspection machines are not connected to the express sorting operation equipment. The image recognition of the security inspection machines relies on manual recognition. There is a lack of security inspection supervision tools. In addition, the image recognition speed of security inspectors is limited, their abilities vary, and they are prone to omissions and mistakes. The security inspection data is not fully collected, and it is not easy to save manually. Summary of the invention

[0003] The purpose of the present invention is to solve the problems in the prior art that the security inspection machine is not connected to the express sorting operation equipment, the image judgment of the security inspection machine relies on manual recognition, there is a lack of security inspection supervision tools, and the security inspectors have limited image judgment speed, uneven abilities and are prone to omissions and errors, security inspection data collection is incomplete, and manual methods are difficult to save. A security inspection and sorting integrated system and method based on AI visual inspection is provided.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] The integrated security inspection and sorting system based on AI visual inspection includes: a conveyor, a security inspection instrument, a single-piece separator, a six-sided barcode scanner, an image acquisition device, a sorting device, and a processing device;

[0006] The conveying device passes through the interior of the single-piece separator, and the conveying device is arranged at one end of the security scanner, and the six-sided code scanner is arranged at the other end of the security scanner, the six-sided code scanner scans the sorted items, and sends the scanned results to the processing device, the image acquisition device is arranged inside the security scanner, the processing device receives the image acquired by the image acquisition device and the scanned results of the six-sided code scanner, and sends a sorting instruction to the sorting device; the sorting device is arranged on one side of the conveying device; the sorting device sorts the items according to the instructions issued by the processing device.

[0007] A further improvement of the present invention is:

[0008] Furthermore, the conveying device includes a first conveyor belt and a second conveyor belt, wherein the first conveyor belt is arranged at one end of the security inspection device; the second conveyor belt is arranged at one end of the six-sided code scanner, and the sorting device is arranged on one side of the second conveyor belt.

[0009] Furthermore, the single-piece separator includes a plurality of mechanical arms, which, under the instruction of the processing unit, separate the packages to prevent the packages from being stacked on top of each other.

[0010] Furthermore, a plurality of scanners are arranged inside the six-sided barcode scanner to comprehensively scan the six sides of the package and read the barcode information of the package.

[0011] Furthermore, after receiving the image captured by the image acquisition device, the processing device processes the image captured by the image acquisition device to detect the target object, and the target detection module includes the following steps:

[0012] Standardize the input image, including image resizing, denoising, pixel value normalization, and color space conversion, to meet the input requirements of the target detection model;

[0013] Extract multi-level features of images based on deep convolutional neural network CNN;

[0014] Based on the feature pyramid network FPN and the path aggregation network PAN, feature maps of different sizes are fused to enhance the detection ability of multi-scale targets;

[0015] The detection head composed of convolutional layers predicts the category probability, target center point coordinates, width, height and confidence of the target box for each pixel position in the fused feature map;

[0016] The prediction results of the detection head are optimized through loss function training; based on multi-classification loss, bounding box regression loss and confidence loss, the training of the target detection model is optimized through weighted combination of multi-task loss functions;

[0017] The detection results are screened, and the overlapping detection frames are removed using the non-maximum suppression algorithm, the frames with high confidence are retained, and the category, location and confidence of the final target are output.

[0018] Furthermore, the multi-task loss function is specifically:

[0019] Confidence loss is VFL Loss

[0020]

[0021] Among them, q is the IOU of the predicted box and the true box, and IOU is the intersection of the predicted box and the true box divided by the union of the two boxes; p is the score, that is, the probability that the predicted box is a positive sample; when q>0, VFL has no hyperparameters for positive samples, that is, there is no attenuation; and when q=0, VFL has hyperparameters for negative samples, γ and α are both hyperparameters, γ is used to reduce the contribution of negative samples; α is used to prevent excessive suppression of negative samples;

[0022] The bounding box regression loss is CIOU+DFL

[0023]

[0024] Among them, IoU is the intersection over union ratio, b and bgt represent the center points of the two rectangular boxes, ρ represents the Euclidean distance between the two rectangular boxes, c represents the diagonal distance of the closed area of ​​the two rectangular boxes, v is used to measure the consistency of the relative proportions of the two rectangular boxes, and α is the weight coefficient;

[0025] DFL(S i , S i+1 )=-((y i+1 -y)log(S i )+(yy i )log(S i+1 ))

[0026] Among them, S i , S i+1 That is, the "prediction value" and "near prediction value" output by the network; y, y i ,y i+1 That is, the "actual value", "label integral value", and "nearby label integral value" of the label.

[0027] Category loss is BCE Loss

[0028]

[0029] Among them, y i represents the true label of category i, 0 or 1, p i represents the predicted probability of category i, and N is the total number of categories;

[0030] The total loss is the weighted sum of the above three parts, and the goal is to optimize bounding box regression, confidence prediction, and category prediction simultaneously;

[0031]

[0032] Among them, λ bbox , obj , cls For different loss weights, represents the bounding box regression loss, represents the confidence loss, The representative class loss is used to adjust the importance of each part.

[0033] Furthermore, the security scanner's shipping port and cargo inlet are provided with lead curtains; there is a gap between the lead curtain and the conveyor belt of the security scanner to prevent packages from stacking; the security scanner's cargo inlet is provided with a high radiation protection baffle to block radiation leakage caused by the gap between the lead curtain and the conveyor belt of the security scanner.

[0034] Furthermore, the processing device receives the image captured by the image acquisition device and the scanning result of the six-sided barcode scanner, and sends a sorting instruction to the sorting device. Specifically, the six-sided barcode scanner sends the read order number to the processing device, the processing device judges the security inspection image captured by the image acquisition device, binds the judgment result with the waybill number, and reports it to the intelligent security inspection platform for order number tracing.

[0035] Furthermore, the sorting device includes a motor and a plurality of telescopic rods; the telescopic rods are connected to the motor; the sorting device sorts the items according to the instructions issued by the processing device, including: the telescopic rods are extended and retracted under the drive of the motor, and different types of items are sorted to the corresponding conveyor belts on one side of the second conveyor belt. When there are suspected dangerous packages, the suspected dangerous packages are sorted to the storage basket on one side of the second conveyor belt; after the dangerous packages are sorted to the storage basket on one side of the second conveyor belt, the order number is scanned by a handheld scanning device, and combined with the X-ray machine image and manual unpacking inspection, whether it is a prohibited item mark is confirmed; if there are dangerous goods, the dangerous goods are sealed and reported to the security department for destruction.

[0036] The integrated security inspection and sorting method based on AI visual detection includes: the first conveyor belt conveys the items to be sorted to the security inspection instrument, the image acquisition device in the security inspection instrument acquires the image of the items to be sorted and transmits it to the processing device, and at the same time, the items to be sorted enter the six-sided barcode scanner, the six-sided barcode scanner performs a comprehensive scan on the six sides of the items to be sorted, reads the information of the package barcode, and uploads it to the processing device, the security inspection image acquired by the image acquisition device is judged, the image judgment result is bound to the waybill number, and reported to the intelligent security inspection platform for number tracing, and the sorting device sorts the items according to the instructions issued by the processing device.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention uses a six-sided barcode scanner to scan the sorted items, and sends the scanned results to a processing device. The image acquisition device is arranged inside the security inspection device. The processing device receives the images acquired by the image acquisition device and the scanned results of the six-sided barcode scanner, and issues a sorting instruction to the sorting device; the sorting device sorts the items according to the instructions issued by the processing device. The present invention uses a processing device to process the order number information and image information acquired by the six-sided barcode scanner and the image acquisition device, automatically and accurately identifies the types of prohibited items, identifies the confidence level, and grades and classifies the prohibited items, which can greatly reduce manpower and misjudgment, reduce the manpower cost of the transfer center, and improve security inspection efficiency. In addition, the security inspection machine of the present invention is unmanned, and the package can be sorted through the sorting line after passing through the security inspection machine, which improves the operating efficiency and reduces the operating process. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 It is a structural diagram of the security inspection instrument;

[0041] Figure 2 This is a schematic diagram of the relationship between the lead curtain and the security scanner;

[0042] Figure 3 It is a schematic diagram of the relationship between the security inspection device and the transmission device;

[0043] Figure 4 A schematic diagram of an image recognized by a processing device. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0045] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0047] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use, it is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0048] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0049] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0050] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0051] See also Figure 1 , Figure 2 , Figure 3 , the present invention discloses an integrated security inspection and sorting system based on AI visual inspection, including: a conveying device, a security inspection instrument, a single-piece separator, a six-sided barcode scanner, an image acquisition device, a sorting device and a processing device;

[0052] The conveying device passes through the interior of the single-piece separator, and the conveying device is arranged at one end of the security scanner, and the six-sided code scanner is arranged at the other end of the security scanner, the six-sided code scanner scans the sorted items, and sends the scanned results to the processing device, the image acquisition device is arranged inside the security scanner, the processing device receives the image acquired by the image acquisition device and the scanned results of the six-sided code scanner, and sends a sorting instruction to the sorting device; the sorting device is arranged on one side of the conveying device; the sorting device sorts the items according to the instructions issued by the processing device.

[0053] The conveying device includes a first conveyor belt and a second conveyor belt, wherein the first conveyor belt is arranged at one end of the security scanner; the second conveyor belt is arranged at one end of the six-sided barcode scanner, and the sorting device is arranged on one side of the second conveyor belt; the belt surface in the security scanner channel remains level with the first conveyor belt and the second conveyor belt.

[0054] The single-piece separator includes a plurality of mechanical arms, which, under the instruction of the processing unit, separate the packages to prevent the packages from being stacked on each other.

[0055] The six-sided barcode scanner has several scanners installed inside to fully scan the six sides of the package and read the barcode information on the package.

[0056] After receiving the image captured by the image acquisition device, the processing device processes the image captured by the image acquisition device to detect the target object. The target detection module includes the following steps:

[0057] Standardize the input image, including image resizing, denoising, pixel value normalization, and color space conversion, to meet the input requirements of the target detection model;

[0058] Extract multi-level features of images based on deep convolutional neural network CNN;

[0059] Based on the feature pyramid network FPN and the path aggregation network PAN, feature maps of different sizes are fused to enhance the detection ability of multi-scale targets;

[0060] The detection head composed of convolutional layers predicts the category probability, target center point coordinates, width, height and confidence of the target box for each pixel position in the fused feature map;

[0061] The prediction results of the detection head are optimized through loss function training; based on multi-classification loss, bounding box regression loss and confidence loss, the training of the target detection model is optimized through weighted combination of multi-task loss functions;

[0062] The detection results are screened, and the overlapping detection frames are removed using the non-maximum suppression algorithm, the frames with high confidence are retained, and the category, location and confidence of the final target are output.

[0063] The multi-task loss function is specifically:

[0064] Confidence loss is VFL Loss

[0065]

[0066] Among them, q is the IOU of the predicted box and the true box, and IOU is the intersection of the predicted box and the true box divided by the union of the two boxes; p is the score, that is, the probability that the predicted box is a positive sample; when q>0, VFL has no hyperparameters for positive samples, that is, there is no attenuation; and when q=0, VFL has hyperparameters for negative samples, γ and α are both hyperparameters, γ is used to reduce the contribution of negative samples; α is used to prevent excessive suppression of negative samples;

[0067] The bounding box regression loss is CIOU+DFL

[0068]

[0069] Among them, IoU is the intersection over union ratio, b and bgt represent the center points of the two rectangular boxes, ρ represents the Euclidean distance between the two rectangular boxes, c represents the diagonal distance of the closed area of ​​the two rectangular boxes, v is used to measure the consistency of the relative proportions of the two rectangular boxes, and α is the weight coefficient;

[0070] DFL(S i , S i+1 )=-((y i+1 -y)log(S i )+(yy i )log(S i+1 ))

[0071] Among them, S i , S i+1 That is, the "prediction value" and "near prediction value" output by the network; y, y i ,y i+1 That is, the "actual value", "label integral value", and "nearby label integral value" of the label.

[0072] Category loss is BCE Loss

[0073]

[0074] Among them, y i represents the true label of category i, 0 or 1, p i represents the predicted probability of category i, and N is the total number of categories;

[0075] The total loss is the weighted sum of the above three parts, and the goal is to optimize bounding box regression, confidence prediction, and category prediction simultaneously;

[0076]

[0077] Among them, λ bbox , obj , cls For different loss weights, represents the bounding box regression loss, represents the confidence loss, Representative class loss is used to adjust the importance of each part

[0078] Lead curtains are installed at the delivery port and the delivery port of the security scanner. There is a gap between the lead curtain and the conveyor belt of the security scanner to prevent the packages from overlapping. A high radiation-proof baffle is installed at the delivery port of the security scanner to block the radiation leakage caused by the gap between the lead curtain and the conveyor belt of the security scanner. The specifications of the high radiation-proof baffle are: 450mm high * 420mm long, with 1mm lead sheet. The baffle is outward-expanding, which will not affect the bag passing and will not interfere with the baffle of the on-site belt conveyor.

[0079] The processing device receives the image captured by the image acquisition device and the scanning result of the six-sided barcode scanner, and sends a sorting instruction to the sorting device. Specifically, the six-sided barcode scanner sends the read order number to the processing device, the processing device judges the security inspection image captured by the image acquisition device, binds the judgment result with the waybill number, and reports it to the intelligent security inspection platform for order number tracing.

[0080] The sorting device includes a motor and a plurality of telescopic rods; the telescopic rods are connected to the motor; the sorting device sorts the items according to the instructions issued by the processing device, including: the telescopic rods are extended and retracted under the drive of the motor, and different types of items are sorted to the corresponding conveyor belt on one side of the second conveyor belt. When there are suspected dangerous packages, the suspected dangerous packages are sorted to the storage basket on one side of the second conveyor belt; after the dangerous packages are sorted to the storage basket on one side of the second conveyor belt, the order number is scanned by a handheld scanning device, and combined with the X-ray machine image and manual unpacking inspection, whether it is a prohibited item is marked and confirmed; if there are dangerous items, the dangerous items are sealed and reported to the security department for destruction.

[0081] At the same time, the security scanner is equipped with a photoelectric detection device, which controls whether the security scanner starts the light source through photoelectric control, thereby reducing the working time of the X-ray machine and the radiation generated.

[0082] A supervision area is set up for the security scanner. The scope of the supervision area is defined as a continuous area around the security scanner where the dose equivalent rate is greater than 2.5μSv / h. The radiation value outside the area is tested regularly.

[0083] The integrated security inspection and sorting method based on AI visual detection includes: a first conveyor belt conveys the items to be sorted to the security inspection instrument, the image acquisition device in the security inspection instrument acquires the image of the items to be sorted and transmits it to the processing device, and at the same time, the items to be sorted enter the six-sided barcode scanner, the six-sided barcode scanner performs a comprehensive scan on the six sides of the items to be sorted, reads the information of the package barcode, and uploads it to the processing device, the security inspection image acquired by the image acquisition device is judged, the image judgment result is bound to the waybill number, and reported to the intelligent security inspection platform for number tracing, and the sorting device sorts the items according to the instructions issued by the processing device.

[0084] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. The integrated security inspection and sorting system based on AI visual inspection is characterized by: include: Conveying device, security inspection device, single piece separator, six-sided barcode scanner, image acquisition device, sorting device and processing device; The conveying device passes through the interior of the single-piece separator, and the conveying device is arranged at one end of the security scanner, and the six-sided code scanner is arranged at the other end of the security scanner, the six-sided code scanner scans the sorted items, and sends the scanned results to the processing device, the image acquisition device is arranged inside the security scanner, the processing device receives the image acquired by the image acquisition device and the scanned results of the six-sided code scanner, and sends a sorting instruction to the sorting device; the sorting device is arranged on one side of the conveying device; the sorting device sorts the items according to the instructions issued by the processing device.

2. The integrated security inspection and sorting system based on AI visual inspection according to claim 1 is characterized in that: The conveying device includes a first conveyor belt and a second conveyor belt, wherein the first conveyor belt is arranged at one end of the security inspection device; the second conveyor belt is arranged at one end of the six-sided code scanner, and the sorting device is arranged at one side of the second conveyor belt.

3. The integrated security inspection and sorting system based on AI visual inspection according to claim 2 is characterized in that: The single-piece separator includes a plurality of mechanical arms, which, under the instruction of the processing unit, separate the packages to prevent the packages from being stacked on each other.

4. The integrated security inspection and sorting system based on AI visual inspection according to claim 3 is characterized in that: The six-sided barcode scanner is internally provided with a plurality of scanners to comprehensively scan the six sides of the package and read the barcode information of the package.

5. The integrated security inspection and sorting system based on AI visual inspection according to claim 4 is characterized in that: After receiving the image acquired by the image acquisition device, the processing device processes the image acquired by the image acquisition device to detect the target object. The target detection module includes the following steps: Standardize the input image, including image resizing, denoising, pixel value normalization, and color space conversion, to meet the input requirements of the target detection model; Extract multi-level features of images based on deep convolutional neural network CNN; Based on the feature pyramid network FPN and the path aggregation network PAN, feature maps of different sizes are fused to enhance the detection ability of multi-scale targets; The detection head composed of convolutional layers predicts the category probability, target center point coordinates, width, height and confidence of the target box for each pixel position in the fused feature map; The prediction results of the detection head are optimized through loss function training. Based on multi-classification loss, bounding box regression loss and confidence loss, the training of the target detection model is optimized through weighted combination of multi-task loss functions. The detection results are screened, and the overlapping detection frames are removed using the non-maximum suppression algorithm, the frames with high confidence are retained, and the category, location and confidence of the final target are output.

6. The integrated security inspection and sorting system based on AI visual inspection according to claim 5 is characterized in that: The multi-task loss function is specifically: Confidence loss is VFL Loss Among them, q is the IOU of the predicted box and the true box, and IOU is the intersection of the predicted box and the true box divided by the union of the two boxes; p is the score, that is, the probability that the predicted box is a positive sample; when q>0, VFL has no hyperparameters for positive samples, that is, there is no attenuation; and when q=0, VFL has hyperparameters for negative samples, γ and α are both hyperparameters, γ is used to reduce the contribution of negative samples; α is used to prevent excessive suppression of negative samples; The bounding box regression loss is CIOU+DFL Among them, IoU is the intersection over union ratio, b and bgt represent the center points of the two rectangular boxes, ρ represents the Euclidean distance between the two rectangular boxes, c represents the diagonal distance of the closed area of ​​the two rectangular boxes, v is used to measure the consistency of the relative proportions of the two rectangular boxes, and α is the weight coefficient; DFL(S i ,S i+1 )=-((y i+1 -y)log(S i )+(y-y i )log(S i+1 )) Among them, S i , S i+1 That is, the "prediction value" and "near prediction value" output by the network; y, y i ,y i+1 That is, the "actual value", "label integral value", and "near label integral value" of the label. Category loss is BCE Loss Among them, y i represents the true label of category i, 0 or 1, p i represents the predicted probability of category i, and N is the total number of categories; The total loss is the weighted sum of the above three parts, and the goal is to optimize bounding box regression, confidence prediction, and category prediction simultaneously; Among them, λ bbox , obj , cls For different loss weights, represents the bounding box regression loss, represents the confidence loss, The representative class loss is used to adjust the importance of each part.

7. The integrated security inspection and sorting system based on AI visual inspection according to claim 6 is characterized in that: The delivery port and the cargo inlet of the security scanner are provided with lead curtains; there is a gap between the lead curtain and the conveyor belt of the security scanner to prevent the packages from overlapping; the cargo inlet of the security scanner is provided with a high radiation protection baffle to block the radiation leakage caused by the gap between the lead curtain and the conveyor belt of the security scanner.

8. The integrated security inspection and sorting system based on AI visual inspection according to claim 7 is characterized in that: The processing device receives the image captured by the image acquisition device and the scanning result of the six-sided barcode scanner, and sends a sorting instruction to the sorting device. Specifically, the six-sided barcode scanner sends the read order number to the processing device, the processing device judges the security inspection image captured by the image acquisition device, binds the judgment result with the waybill number, and reports it to the intelligent security inspection platform for order number tracing.

9. The integrated security inspection and sorting system based on AI visual detection according to claim 8 is characterized in that: The sorting device includes a motor and a plurality of telescopic rods; the telescopic rods are connected to the motor; the sorting device sorts the items according to the instructions issued by the processing device, including: the telescopic rods are extended and retracted under the drive of the motor, and different types of items are sorted to the corresponding conveyor belts on one side of the second conveyor belt. When there are suspected dangerous packages, the suspected dangerous packages are sorted to the storage basket on one side of the second conveyor belt; after the dangerous packages are sorted to the storage basket on one side of the second conveyor belt, the order number is scanned by a handheld scanning device, and combined with the X-ray machine image and manual unpacking inspection, whether it is a prohibited item is marked and confirmed; if there are dangerous items, the dangerous items are sealed and reported to the security department for destruction.

10. The integrated security inspection and sorting method based on AI visual inspection is characterized by: The integrated security inspection and sorting system based on AI visual inspection as described in claim 9 is adopted, and the method includes: the first conveyor belt conveys the items to be sorted to the security inspection instrument, the image acquisition device in the security inspection instrument collects the image of the items to be sorted and transmits it to the processing device, and at the same time, the items to be sorted enter the six-sided barcode scanner, the six-sided barcode scanner comprehensively scans the six sides of the items to be sorted, reads the information of the package barcode, and uploads it to the processing device, the security inspection image collected by the image acquisition device is judged, the judgment result is bound to the waybill number, and reported to the intelligent security inspection platform for number tracing, and the sorting device sorts the items according to the instructions issued by the processing device.

Images