A sealing and locking device and method based on AI vision

Abstract

The present application provides an AI-vision-based sealing and locking device and method. The AI-vision-based sealing and locking device comprises: a posture adjustment section, a sealing and locking section, a camera, and a processing unit. The posture adjustment section comprises a first conveying device, a support beam, and a pair of clamping arm assemblies. The support beam is disposed above the first conveying device. The clamping arm assembly comprises a movable clamping arm and a side belt conveyor. The camera is located above the first conveying device and is used to capture luggage images. The processing unit is used to perform real-time deflection angle analysis on the luggage images. While the clamping arm assembly is gripping the luggage, the transmission differential speed of the two side belt conveyors is controlled to adjust the luggage according to the luggage's deflection angle. The AI-vision-based sealing and locking device and method of the present application performs real-time deflection angle analysis on the luggage images, controls the transmission differential speed of the two side belt conveyors to adjust the luggage's deflection angle according to the luggage's deflection angle, and improves the quality of sealing and locking.

Description

Technical Field

[0001] The present application belongs to the technical field of luggage sealing and locking, and more specifically, to a sealing and locking device and method based on AI vision. Background Art

[0002] A lock system is a device for monitoring risky luggage. For example, the utility model patent with publication number CN218974925U discloses that the lock system is equipped with a Bluetooth main control chip, a Bluetooth onboard antenna, an acceleration sensor, an unlocking motor, an NFC sensing module, a gesture sensor and other devices for tracking luggage, status detection, and anti-tampering detection.

[0003] To improve sealing and locking efficiency, the existing technology uses an automated sealing and locking mechanism to seal risky luggage. Before the luggage is sealed and locked, the sealing and locking mechanism uses a swing arm and other structures to mechanically correct the direction of the luggage. However, since the posture and position of the luggage during transportation are uncertain, for example, if the luggage is not centered during transportation, is offset at a large angle, is placed horizontally, or is facing up, the mechanical action may not be able to correct the luggage properly. The correction result may have large errors, and it is impossible to achieve multi-dimensional standardization, which affects the quality of sealing and locking. Summary of the Invention

[0004] The purpose of the embodiments of the present application is to provide an AI vision-based sealing and locking device and method to solve the technical problem of insufficient quality in the luggage sealing process in the prior art.

[0005] To achieve the above objectives, the technical solution adopted in this application is to provide an AI vision-based sealing and locking device, including: a posture adjustment section, a sealing and locking section, a camera, and a processing unit;

[0006] The posture adjustment section includes a first conveying device, a supporting beam, and a pair of clamping arm assemblies, wherein the supporting beam is arranged above the first conveying device, and the clamping arm assembly includes a movable clamping arm and a side belt conveyor, wherein the pair of movable clamping arms are movably arranged on the supporting beam, and the side belt conveyors are arranged on the corresponding movable clamping arms, and the pair of side belt conveyors are arranged parallel to each other and perpendicular to the first conveying device;

[0007] The camera is located above the first conveying device and is used to capture images of the luggage. The processing unit is used to perform real-time deflection angle analysis on the luggage images. When the clamping arm assembly is clamping the luggage, the transmission differential speed of the two side belt conveyors is controlled according to the deflection angle of the luggage to adjust the luggage.

[0008] The sealing and locking section includes a second conveying device and a sealing and locking mechanism. The second conveying device is arranged in parallel with the first conveying device. The sealing and locking mechanism is located on the second conveying device and is used to seal and lock the luggage after the deflection angle is adjusted.

[0009] In a preferred embodiment, the movable clamping arm is provided with an elastic guide mechanism and a photoelectric sensing unit, the side belt conveyor is provided with a detection piece, the side belt conveyor is slidingly connected to the elastic guide mechanism, and the detection piece is correspondingly provided to the photoelectric sensing unit.

[0010] In a preferred embodiment, a ball screw linear module and a power unit are provided on the support beam, the power unit is connected to the ball screw linear module, and a pair of clamping arm assemblies are connected to the ball screw linear module.

[0011] In a preferred embodiment, the second conveying device includes a conveying device body, a flip shaft, a seat bearing and a flip power unit, the conveying device body is arranged on the flip shaft, the flip shaft is arranged in the seat bearing, and the flip power unit is connected to the flip shaft.

[0012] The present application also provides an AI vision-based sealing and locking method, based on the above-mentioned AI vision-based sealing and locking device, comprising the steps of:

[0013] Acquire baggage images in real time;

[0014] Perform real-time deflection angle α analysis on baggage images;

[0015] According to the luggage deflection angle α, the transmission differential of the two side belt conveyors is controlled to adjust the luggage;

[0016] The luggage after the deflection angle is adjusted is sealed and locked.

[0017] In a preferred embodiment, the method for performing real-time deflection angle analysis on a baggage image comprises the steps of:

[0018] Acquire baggage targets in real time based on baggage images;

[0019] Get the minimum bounding rectangle of the baggage target;

[0020] The deflection angle α is calculated using the minimum rectangular box.

[0021] In a preferred embodiment, the method for calculating the deflection angle α of the minimum rectangular frame includes the steps of:

[0022] The baggage transfer direction is the reference direction;

[0023] Detect the end features of the baggage object and determine the top and bottom of the baggage object;

[0024] The deflection angle α in the minimum circumscribed rectangular frame vector state is determined according to the relative positions of the top and the bottom.

[0025] In a preferred embodiment, a method for adjusting luggage by controlling the transmission differential of two side belt conveyors according to the luggage deflection angle α comprises the following steps:

[0026] Detect the front and back of the luggage target. When the front of the luggage is facing up, output A=0, and when the back of the luggage is facing up, output A=1;

[0027] The deflection angle α is normalized so that α∈(-π,π];

[0028] The angle of the luggage rotation caused by controlling the transmission differential of the two side belt conveyors is β, where

[0029] .

[0030] In a preferred embodiment, the minimum circumscribed rectangular frame of the baggage object is obtained, the width change of the minimum circumscribed rectangular frame is detected in real time, and the distance between the pair of side belt conveyors is dynamically adjusted so that the width of the minimum circumscribed rectangular frame and the distance between the pair of side belt conveyors maintain synchronous changes.

[0031] In a preferred embodiment, the real-time transmission speed V1 of the first conveying device is obtained, and the standard rated differential speed V2 and the compensated rated differential speed V3 are set; the transmission speeds of the two side belt conveyors are speed1 and speed2 respectively, where

[0032] .

[0033] In a preferred embodiment, if the output A=1, after adjusting the angle of the luggage, the following steps are further included:

[0034] When the luggage is transferred to the second conveyor after the angle adjustment, the second conveyor performs a rotation operation greater than 90 degrees to turn the luggage over.

[0035] The beneficial effects of the AI vision-based sealing and locking device and method provided in this application are: compared with the existing technology, real-time deflection angle analysis of luggage images is performed, and the transmission differential speed of the two side belt conveyors is controlled according to the luggage deflection angle to adjust the luggage deflection angle, so that the luggage is corrected, which can not only protect the luggage but also improve the quality of sealing and locking. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0037] Figure 1 A schematic diagram of the three-dimensional structure of the AI vision-based sealing and locking device provided in an embodiment of the present application;

[0038] Figure 2 for Figure 1 A schematic diagram of the three-dimensional structure of the main part of the posture adjustment section;

[0039] Figure 3 for Figure 2 A schematic diagram of the three-dimensional structure of the clamp arm assembly;

[0040] Figure 4 for Figure 3 Magnified view of area B in FIG;

[0041] Figure 5 A top view of the posture adjustment section when the luggage just enters the first conveyor device;

[0042] Figure 6 A top view of the posture adjustment section of a pair of side conveyors just as they come into contact with the luggage;

[0043] Figure 7 It is a top view of the posture adjustment section when the luggage is adjusted into place;

[0044] Figure 8 It is a top view of the posture adjustment section of a pair of side belt conveyors after reset;

[0045] Figure 9 A schematic diagram of the three-dimensional structure of the second conveying device provided in an embodiment of the present application;

[0046] Figure 10 This is a schematic diagram of luggage rotation when A=0 in the AI vision-based sealing and locking method provided in an embodiment of the present application;

[0047] Figure 11 Schematic diagram of luggage rotation when A=1 in the AI vision-based sealing and locking method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to make the technical problems, technical solutions and beneficial effects to be solved by this application more clearly understood, this application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0049] It should be noted that when an element is referred to as being “fixed on” or “disposed on” another element, it may be directly on the other element or indirectly on the other element. When an element is referred to as being “connected to” another element, it may be directly connected to the other element or indirectly connected to the other element.

[0050] It should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do 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 this application.

[0051] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. Throughout the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0052] Please also refer to Figures 1 to 4 The AI vision-based sealing and locking device 100 provided in an embodiment of the present application is now described. The AI vision-based sealing and locking device 100 includes: a posture adjustment section 10, a sealing and locking section 20, a camera 30, and a processing unit (not shown).

[0053] Specifically, the posture adjustment section 10 includes a first conveying device 11, a supporting beam 12 and a pair of clamping arm assemblies 13. The supporting beam 12 is arranged above the first conveying device 11. The clamping arm assembly 13 includes a movable clamping arm 131 and a side belt conveyor 132. The pair of movable clamping arms 131 are movably arranged on the supporting beam 12. The side belt conveyor 132 is arranged on the corresponding movable clamping arms 131. The pair of side belt conveyors 132 are parallel to each other and perpendicular to the first conveying device 11.

[0054] The camera 30 is located above the first conveying device 11 and is used to capture luggage images. The processing unit is used to perform real-time deflection angle analysis on the luggage images. When the clamping arm assembly 13 clamps the luggage, the transmission differential of the two side belt conveyors 132 is controlled according to the luggage deflection angle to adjust the luggage.

[0055] Specifically, the sealing and locking section 20 includes a second conveying device 21 and a sealing and locking mechanism 22. The second conveying device 21 is arranged in parallel with the first conveying device 11. The sealing and locking mechanism 22 is located on the second conveying device 21 and is used to seal and lock the luggage after the deflection angle is adjusted.

[0056] It is understood that the first conveying device 11 and the second conveying device 21 can be conveyor belts, stick conveying devices or universal wheel conveying devices, so as to stably convey the luggage. The two ends of the support beam 12 can be fixed above the first conveying device 11 by vertical support columns.

[0057] In one embodiment, the second conveying device 21 and the first conveying device 11 are both provided with a plurality of infrared sensors 211 , and the infrared sensors 211 are used to detect the transfer position or speed of the luggage.

[0058] In one embodiment, the processing unit is a central processing unit, or a remotely connected backend server, which is capable of processing and analyzing the luggage image and sending control commands.

[0059] In one embodiment, the AI vision-based sealing and locking device 100 further includes a buffer section 40 , which is connected to the second conveying device 21 and is used to buffer the sealed and locked luggage.

[0060] The luggage image can be a standard top-down image of the luggage, or a relative top-down image with a certain offset relative to the standard top-down image, so as to reflect the overall deflection angle α of the luggage through the luggage image. Figures 5 to 8 shown.

[0061] Thus, the AI vision-based sealing and locking device 100 provided in this application offers the following advantages: Compared to existing technologies, when luggage enters the first conveyor 11 and stops moving, the luggage's deflection angle is adjusted by controlling the transmission differential of the two side conveyors 132 based on the luggage's deflection angle, thereby aligning the luggage. Specifically, the two side conveyors 132 can be controlled to rotate in opposite directions to create a large speed difference, applying opposite frictional forces on both sides of the luggage. This prevents squeezing of the luggage during the alignment process, thereby protecting the luggage. This improves the quality of the sealing and locking process.

[0062] In one embodiment, see Figure 4 and Figure 5The movable clamping arm 131 is provided with an elastic guide mechanism 133 and a photoelectric sensing unit 134, and the side belt conveyor 132 is provided with a detection piece 135. The side belt conveyor 132 is slidingly connected to the elastic guide mechanism 133, and the detection piece 135 is correspondingly arranged to the photoelectric sensing unit 134.

[0063] It can be understood that the elastic guide mechanism 133 can be a mechanism composed of a guide rail and a spring, or a mechanism composed of a guide rod and a spring. The purpose is that when the side belt conveyor 132 clamps the luggage, the side belt conveyor 132 can move backward under the guidance of the elastic guide mechanism 133. When the pressure reaches a certain level, the detection piece 135 triggers the photoelectric sensing unit 134, so that the pair of clamping arm assemblies 13 stop moving closer, which is conducive to maintaining a relatively constant pressure on the luggage by the pair of side belt conveyors 132.

[0064] In one embodiment, see Figure 2 A ball screw linear module 121 and a power unit 122 are provided on the support beam 12 . The power unit 122 is connected to the ball screw linear module 121 , and a pair of clamping arm assemblies 13 are connected to the ball screw linear module 121 .

[0065] It can be understood that the power unit 122 includes a servo motor or a stepper motor, and of course can also include a reducer connected to the motor. The ball screw linear module 121 is bidirectionally driven, that is, the power unit 122 can simultaneously control a pair of clamping arm assemblies 13 to move closer to or away from each other, so as to achieve coordinated cooperation.

[0066] In one embodiment, see Figure 9 The second conveying device 21 includes a conveying device body 212, a flip shaft 213, a seat bearing 214 and a flip power unit (not shown in the figure). The conveying device body 212 is arranged on the flip shaft 213, the flip shaft 213 is arranged in the seat bearing 214, and the flip power unit is connected to the flip shaft 213.

[0067] It is understandable that the turning power unit can drive the conveying device body 212 to turn toward the side close to the first conveying device 11, so that when the sealing and locking mechanism is damaged, it is convenient to repair. The turning power unit can be a servo motor or a stepping motor.

[0068] The present application further provides an AI vision-based sealing and locking method, based on the AI vision-based sealing and locking device 100 as described above, the AI vision-based sealing and locking method includes the following steps:

[0069] Acquire baggage images in real time;

[0070] Perform real-time deflection angle α analysis on baggage images;

[0071] Control the transmission differential of the two side belt conveyors 132 to adjust the luggage according to the luggage deflection angle α;

[0072] The luggage after the deflection angle is adjusted is sealed and locked.

[0073] It is understandable that the luggage image may be a standard top-down image of the luggage, or a relative top-down image with a certain offset relative to the standard top-down image, so as to reflect the overall deflection angle α of the luggage through the luggage image.

[0074] In one embodiment, after the baggage enters the first conveyor 11 and stops moving, the luggage deflection angle α is adjusted by controlling the transmission differential of the two side conveyors 132 based on the luggage deflection angle α, thereby aligning the baggage. Specifically, the two side conveyors 132 can be controlled to rotate in opposite directions to create a large speed difference, thereby applying opposite friction forces on both sides of the baggage.

[0075] In another embodiment of the present application, a method for performing real-time deflection angle analysis on a baggage image includes the steps of:

[0076] Acquire baggage targets in real time based on baggage images;

[0077] Get the minimum bounding rectangle of the baggage target;

[0078] The deflection angle α is calculated using the minimum rectangular box.

[0079] It is understood that the baggage target can be obtained based on a deep learning target detection model or detected by image difference method. When obtaining the minimum bounding rectangle of the baggage target, the direction of the baggage target can be adapted, and then the deflection angle α can be calculated by converting it into a geometric problem.

[0080] In another embodiment of the present application, a method for calculating a deflection angle α of a minimum rectangular frame includes the steps of:

[0081] The baggage transfer direction is the reference direction;

[0082] Detect the end features of the baggage object and determine the top and bottom of the baggage object;

[0083] The deflection angle α in the minimum circumscribed rectangular frame vector state is determined according to the relative positions of the top and the bottom.

[0084] It is understandable that the end features should be understood as the bottom features and top features of the luggage. For example, the bottom features include wheels, and the top features include handles and pull rods. In this way, deep learning target detection can be used to determine the top and bottom ends of the luggage target. In this way, after obtaining the minimum circumscribed rectangular frame, the direction of the luggage can also be determined to obtain the deflection angle α in the vector state.

[0085] For further information, please also refer to Figure 9 and Figure 10 The method for adjusting the luggage by controlling the transmission differential of the two side belt conveyors 132 according to the luggage deflection angle α comprises the following steps:

[0086] Detect the front and back of the luggage target. When the front of the luggage is facing up, output A=0, and when the back of the luggage is facing up, output A=1;

[0087] The deflection angle α is normalized so that α∈(-π,π];

[0088] The angle of the luggage rotation caused by controlling the transmission differential of the two side belt conveyors 132 is β, where

[0089] .

[0090] It's understood that, in addition to obtaining the deflection angle α in the vector state, the front and back sides of the luggage object are also detected. This allows for simultaneous determination of both the deflection angle α and the front and back side luggage posture parameters. After normalizing the deflection angle α, the above formula directly adjusts the angle for all possible luggage postures, ensuring that the adjusted opening and closing sides of the luggage remain on the same side when the rotation angle is minimized.

[0091] In this way, the effect of combining with the sealing and locking section 20 is: when the adjusted luggage reaches the second conveying device 21, the sealing and locking mechanism 22 of the sealing and locking section 20 is used to seal and lock the adjusted luggage, which can ensure that the locking position is uniformly located on the opening and closing side of the luggage box, thereby improving the automation standard level of the sealing and locking process, facilitating the implementation of automatic unlocking, identification, and detection, and on the other hand, the anti-dismantling detection mechanism of the sealing lock can be fully utilized. For example, a customs sealing control system with publication number CN218974925U describes that when a passenger forcibly and illegally opens the luggage, the gesture sensor at the bottom of the sealing lock can sense the object, and the sensing information of the gesture sensor can be sent to the background through the Bluetooth on-board antenna, so as to prompt the staff to stop the passenger's behavior in time.

[0092] It is worth mentioning that, in this embodiment, the luggage transfer direction is used as the reference direction. When β>0, it can be understood as clockwise rotation, and when β<0, it can be understood as counterclockwise rotation.

[0093] In another embodiment of the present application, when ︱β︱>90°, the rotation angle of the luggage is relatively large. In order to utilize the side belt conveyor 132 to complete the entire luggage rotation process, a pair of side belt conveyors 132 need to undergo a process of first separation and then clamping, rather than a continuous clamping process. However, during the separation process, the side belt conveyors 132 may experience virtual rotation and stalling. The virtual rotation means that the side belt conveyor 132 does not form effective contact with the luggage, and there is no friction or sliding friction. Only the side belt conveyor 132 rotates, but the luggage does not rotate. The stall means that the pressure between the side belt conveyor 132 and the luggage is too high, and there is insufficient space for the luggage to rotate, resulting in sliding friction, causing wear and extrusion on the luggage.

[0094] Please also refer to Figure 5 To solve the above problem, the minimum circumscribed rectangular frame of the baggage target is obtained, the width change of the minimum circumscribed rectangular frame is detected in real time, and the distance between the pair of side belt conveyors 132 is dynamically adjusted so that the width of the minimum circumscribed rectangular frame and the distance between the pair of side belt conveyors 132 keep changing synchronously.

[0095] It should be understood that the minimum circumscribed rectangular frame should be understood by those skilled in the art as distinct from the aforementioned minimum circumscribed rectangular frame. In this embodiment, the minimum circumscribed rectangular frame does not rotate with the luggage's rotation. In this case, the width of the minimum circumscribed rectangular frame should be understood as the length perpendicular to the luggage's conveying direction. After the pair of side conveyors 132 begin rotating the luggage, the changes in the width of the minimum circumscribed rectangular frame are monitored in real time, and the spacing between the pair of side conveyors 132 is dynamically adjusted to ensure that the width of the minimum circumscribed rectangular frame and the spacing between the pair of side conveyors 132 change synchronously.

[0096] In this way, during this process, the spacing between a pair of side belt conveyors 132 can be changed to dynamically adapt to the changes in the luggage's posture in real time, providing just the right space for the luggage to rotate, while also continuously providing the static friction required for the luggage's rotation to protect the integrity of the luggage.

[0097] In order to further adapt the side belt conveyor 132 to the changes in the luggage's posture and improve the sealing and locking efficiency, please refer to Figure 6 , obtain the real-time transmission speed V1 of the first conveying device 11, and set the standard rated differential speed V2 and the compensated rated differential speed V3; the transmission speeds of the two side belt conveyors 132 are speed1 and speed2 respectively, where

[0098] .

[0099] It can be understood that, in this embodiment, the transmission speed of each side belt conveyor 132 is divided into three dimensions for control, and the actual transmission speed of each side belt conveyor 132 is affected by the speed of the three dimensions, and the influence relationship is different.

[0100] Regarding the first dimension, the real-time transmission speed V1 is determined by the first conveying device 11 , that is, when the two side belt conveyors 132 move at the real-time transmission speed V1 , there is pressure between the two side belt conveyors 132 and the luggage, but the friction is zero.

[0101] Regarding the second dimension, assuming the first conveyor 11 is stationary and the pressure is constant during luggage posture adjustment, the speeds of the two side conveyors 132, V2 and -V2, respectively, are sufficient to maintain luggage rotation using static friction. The standard rated differential speed V2 can be set based on the typical coefficient of friction of luggage surfaces. In a preferred embodiment, the speed difference between the two side conveyors 132 is gradually increased until the luggage's rotational speed reaches its maximum speed (i.e., the critical value at which static friction transitions to sliding friction). This speed is then used as the standard rated differential speed V2, maximizing efficiency while also accommodating luggage with varying friction coefficients.

[0102] However, under test conditions, when combining the first dimension's real-time transmission speed V1 with the second dimension's standard rated differential speed V2, the width change rate of the minimum circumscribed rectangular frame fluctuated during rotation, as luggage is typically rectangular. This made it difficult to maintain synchronization when controlling the movement speed of the clamping arm assembly 13 using visual inspection. During luggage adjustments, the side belt conveyor 132 experienced false rotation and stalling, preventing the theoretical superposition effect from being achieved. To overcome this integration obstacle, a third dimension was ultimately introduced: V3×|sinβ|, which compensates for speed differences, maintaining a relatively constant width change rate during rotation.

[0103] In this way, when the pair of side conveyors 132 respectively operate at the transmission speeds speed1 and speed2, which are the sum of three dimensions, the following objectives can be achieved: while the luggage is being conveyed, the pair of side conveyors 132 continuously exert two opposite static friction forces on the luggage to adjust its posture. Ultimately, this achieves the beneficial effects of protecting the integrity of the luggage and improving the efficiency of sealing and locking.

[0104] Optimally, after the luggage angle is adjusted to the correct position, speed1=speed2=V1, and a pair of side conveyors remain stationary, maintaining a clamping state to convey the luggage to the second conveyor to ensure that no deflection occurs during the process of entering the second conveyor.

[0105] In all the above embodiments, the speed difference of the pair of side belt conveyors 132 of the clamping arm assembly 13 is used to adjust the luggage's posture. However, the front and back sides of the luggage cannot be adjusted. In order to further standardize the luggage's posture, in another embodiment of this application, please refer to Figure 7 If the output A=1, after adjusting the angle of the luggage, the following steps are also included:

[0106] When the luggage is transferred to the second conveying device 21 after the angle adjustment, the second conveying device 21 rotates more than 90 degrees to turn the luggage over.

[0107] It is understood that if the output A = 1, it indicates that the luggage is facing backward. When the luggage is transferred to the second conveyor 21, the second conveyor 21's flipping function can be used to flip the luggage over, thereby achieving simultaneous adjustment of the rotation angle and the front and back orientation. In a preferred embodiment, the flipped luggage re-enters the first conveyor 11, where the posture adjustment section 10 performs a rotation adjustment on the luggage.

[0108] This further improves the automation standard of the sealing and locking process. For example, after each piece of luggage is sealed and locked, its sealing position, locking position, and locking direction remain consistent, which is conducive to the implementation of subsequent processes such as automated unlocking, identification, and detection.

[0109] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A sealing and locking method based on AI vision, characterized in that: An AI vision-based sealing and locking device, comprising: a posture adjustment section, a sealing and locking section, a camera, and a processing unit; The posture adjustment section includes a first conveying device, a supporting beam, and a pair of clamping arm assemblies, wherein the supporting beam is arranged above the first conveying device, and the clamping arm assembly includes a movable clamping arm and a side belt conveyor, wherein the pair of movable clamping arms are movably arranged on the supporting beam, and the side belt conveyors are arranged on the corresponding movable clamping arms, and the pair of side belt conveyors are arranged parallel to each other and perpendicular to the first conveying device; The camera is located above the first conveying device and is used to capture images of the luggage. The processing unit is used to perform real-time deflection angle analysis on the luggage images. When the clamping arm assembly is clamping the luggage, the transmission differential speed of the two side belt conveyors is controlled according to the deflection angle of the luggage to adjust the luggage. The AI vision-based sealing and locking method includes the following steps: Acquire baggage images in real time; Perform real-time deflection angle α analysis on baggage images; According to the luggage deflection angle α, the transmission differential of the two side belt conveyors is controlled to adjust the luggage; Seal and lock the luggage after the deflection angle is adjusted; The method for adjusting luggage by controlling the transmission differential of two side belt conveyors according to the luggage deflection angle α comprises the following steps: Detect the front and back of the luggage target. When the front of the luggage is facing up, output A=0, and when the back of the luggage is facing up, output A=1; The deflection angle α is normalized so that α∈(-π,π]; The angle of the luggage rotation caused by controlling the transmission differential of the two side belt conveyors is β, where 。 2. The AI vision-based sealing and locking method according to claim 1, characterized in that: The sealing and locking section includes a second conveying device and a sealing and locking mechanism. The second conveying device is arranged in parallel with the first conveying device. The sealing and locking mechanism is located on the second conveying device and is used to seal and lock the luggage after the deflection angle is adjusted.

3. The AI vision-based sealing and locking method according to claim 1, characterized in that: The movable clamping arm is provided with an elastic guide mechanism and a photoelectric sensing unit, the side belt conveyor is provided with a detection piece, the side belt conveyor is slidably connected to the elastic guide mechanism, and the detection piece is correspondingly provided to the photoelectric sensing unit.

4. The AI vision-based sealing and locking method according to claim 1, wherein: A ball screw linear module and a power unit are provided on the supporting crossbeam. The power unit is connected to the ball screw linear module. A pair of clamping arm assemblies are connected to the ball screw linear module.

5. The AI vision-based sealing and locking method according to claim 2, characterized in that: The second conveying device includes a conveying device body, a flip shaft, a seat bearing and a flip power unit. The conveying device body is arranged on the flip shaft, the flip shaft is arranged in the seat bearing, and the flip power unit is connected to the flip shaft.

6. The AI vision-based sealing and locking method according to claim 1, characterized in that: Performing real-time deflection angle analysis on baggage images includes the following steps: Acquire baggage targets in real time based on baggage images; Get the minimum bounding rectangle of the baggage target; The deflection angle α is calculated using the minimum rectangular box.

7. The AI vision-based sealing and locking method according to claim 6, characterized in that: The deflection angle α is calculated using the minimum rectangular frame, including the following steps: The baggage transfer direction is the reference direction; Detect the end features of the baggage object and determine the top and bottom of the baggage object; The deflection angle α in the minimum circumscribed rectangular frame vector state is determined according to the relative positions of the top and the bottom.

8. The AI vision-based sealing and locking method according to claim 7, characterized in that: Obtain the minimum circumscribed rectangular frame of the baggage target, detect the width change of the minimum circumscribed rectangular frame in real time, and dynamically adjust the distance between a pair of side belt conveyors so that the width of the minimum circumscribed rectangular frame and the distance between the pair of side belt conveyors keep changing synchronously.

9. The AI vision-based sealing and locking method according to claim 8, characterized in that: Get the real-time transmission speed V1 of the first conveyor device, and set the standard rated differential speed V2 and the compensated rated differential speed V3; the transmission speeds of the two side belt conveyors are speed1 and speed2 respectively, where 。

Images