An integrated system and method for intelligent inspection of boxed cargo

Abstract

The present invention discloses an integrated intelligent inspection system and method for box-type cargo, the integrated intelligent inspection system for box-type cargo comprises a central control platform, a loading and unloading robot, a conveying device, an inspection device, a mark recognition device, and a separation device. The present invention solves the problem that the robots or automated equipment for logistics loading and unloading of customs in the prior art cannot realize the processes of cargo identification, quantity statistics, and abnormal situation screening, thereby leading to a long customs clearance time for cargo through the integrated intelligent inspection system for box-type cargo.

Description

Technical Field

[0001] The present invention relates to the field of intelligent devices, and in particular to an integrated system and method for intelligent inspection of boxed cargo. Background Art

[0002] At present, there are prominent problems in the inspection of imported cold chain goods, such as poor working environment, low manual efficiency, and too many management personnel. For the logistics loading and unloading of customs, existing technologies have introduced some robots or automated equipment, but these loading and unloading robots and automated equipment only stay at the loading and unloading level, and do not open up the entire process of customs import and export inspection. They cannot realize the process of identifying goods, counting quantities, and identifying abnormal situations, and thus cannot assist staff in inspection work, resulting in long customs clearance time for goods, and the goods cannot be quickly inspected and cleared.

[0003] Therefore, the existing technology still needs to be improved and developed. Summary of the invention

[0004] The main purpose of the present invention is to provide an integrated system and method for intelligent inspection of box-type cargo, aiming to solve the problem that robots or automated equipment for logistics loading and unloading of customs in the prior art are unable to realize processes such as identification, quantity counting and abnormal situation screening of cargo, thus resulting in long customs clearance time for cargo.

[0005] In order to achieve the above-mentioned object, the first aspect of the present invention provides an integrated intelligent inspection system for box-type cargo, wherein the integrated intelligent inspection system for box-type cargo includes a central control platform, a loading and unloading robot, a conveying device, an inspection device, a mark recognition device and a separation device;

[0006] The central control platform is connected to the loading and unloading robot, the conveying device, the inspection device, the mark recognition device and the separation device through a network, and is used to obtain inspection instructions, parse the inspection instructions to generate execution action instructions, and send the execution actions to the loading and unloading robot. At the same time, according to the customs declaration information obtained in advance, the risk attribute quantitative parameters of the target goods are generated, and the risk attribute quantitative parameters are displayed on the central control platform;

[0007] The loading and unloading robot is used to receive an execution action instruction, grab the target goods according to the execution action instruction, and place the target goods on the conveying device;

[0008] The conveying device is used to transport the target goods to the inspection device;

[0009] The inspection device is connected to the end of the conveying device, and is used to measure the size, quantity and weight of the target goods, obtain a measurement result, compare the preset parameter information with the measurement result, obtain an inspection result, judge whether the target goods are abnormal according to the inspection result, obtain a first judgment result, upload the first judgment result to the central control platform, and convey the target goods to the mark recognition device;

[0010] The mark recognition device is connected to the end of the inspection device, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device;

[0011] The separation device is connected to the end of the mark recognition device, and is used to receive the judgment result generated by the central control platform based on the first judgment result and the second judgment result, transport the target goods to the target location according to the judgment result, and push the target location information to the central control platform.

[0012] Optionally, the loading and unloading robot includes a chassis, a vehicle body control box, a hybrid robotic arm, a grasping end, a visual perception module and a control module;

[0013] The chassis is located at the bottom of the loading and unloading robot and is used to drive the loading and unloading robot to move;

[0014] The vehicle body control box is fixedly arranged on the chassis and is used to load the control module and the visual perception module;

[0015] The hybrid robot arm is arranged on the chassis and is used to drive the grasping end to move in space;

[0016] The grabbing end is arranged at the end of the hybrid robot arm, and is used to absorb and transfer the target goods;

[0017] The visual perception module is arranged on the hybrid robotic arm, and is used to obtain the grasping posture of the target cargo;

[0018] The control module is arranged on the vehicle body control box, and is used to control the chassis, the hybrid mechanical arm and the grabbing end to absorb the target cargo and then transfer it to the conveying device;

[0019] Wherein, the control module is electrically connected to the chassis, the visual perception module, the hybrid robotic arm and the grasping end respectively.

[0020] Optionally, the conveying device includes a frame, a back-rolling conveyor belt, a lifting module and a roller conveyor belt;

[0021] The frame is used to fix the back-rolling conveyor belt, the lifting module and the roller conveyor belt;

[0022] The back roller conveyor belt and the lifting module are slidably arranged on the frame, the roller conveyor belt is detachably arranged on the frame, the roller conveyor belt and the lifting module are arranged on the same side of the frame, wherein the back roller conveyor belt is arranged on the opposite side of the roller conveyor belt and the lifting module;

[0023] The back-rolling conveyor belt is used to place the target goods;

[0024] The lifting module is used to adjust the height of the back-rolling conveyor belt to the height of the roller conveyor belt;

[0025] The roller conveyor belt is used to convey the target goods to the inspection device.

[0026] Optionally, the inspection device includes a conveying module, a weighing module, a 3D camera, a control cabinet and an audible and visual alarm module;

[0027] The conveying module is used to transfer the target goods to the weighing module;

[0028] The weighing module is disposed below the conveying module, and is used to measure the weight of the target goods, obtain weighing information, and transmit the weighing information to the control cabinet;

[0029] The 3D camera is fixedly mounted on the control cabinet, and is used to measure the size and count the quantity of the target goods, obtain size information and quantity information, and transmit the size information and quantity information to the control cabinet;

[0030] The control cabinet is arranged beside the conveying module, and is used to obtain the measurement result according to the weighing information, the quantity information and the size information, compare the parameter information with the measurement result to obtain the inspection result, judge whether the target goods are abnormal according to the inspection result, obtain a first judgment result, and upload the first judgment result to the central control platform;

[0031] The sound and light alarm module is arranged on the control cabinet and is used to determine whether to give an alarm according to the inspection result;

[0032] Wherein, the control cabinet is electrically connected to the conveying module, the weighing module, the 3D camera and the sound and light alarm module respectively.

[0033] Optionally, the mark recognition device includes a mark recognition rack, an industrial camera, a motion mechanism, a rotating platform and a mark recognition conveying module;

[0034] The mark recognition frame is used to fix the rotating platform and the motion mechanism;

[0035] The motion mechanism is arranged on the mark recognition frame and is used to rotate the target goods;

[0036] The industrial camera is used to take an initial photograph of the target goods to obtain an initial position photograph, control the motion mechanism to move to the mark position according to the initial position photograph, take a photograph of the mark of the target goods to obtain a mark position photograph, extract the mark information of the target goods according to the mark position photograph, compare the pre-acquired customs declaration information with the mark information, judge whether the target goods are abnormal, obtain a second judgment result, and upload the second judgment result and the mark information to the central control platform;

[0037] The movement mechanism is arranged on the mark recognition frame and is used to move according to the initial position photo;

[0038] The mark recognition and conveying module is arranged on the mark recognition frame, and is used for conveying the target goods to the separation device.

[0039] Optionally, the separation device includes a roller conveyor belt, a rotating mechanism and an information transmission device;

[0040] The roller conveyor belt is connected to the rotating mechanism and is used to convey the target goods to the target location;

[0041] The rotating mechanism is used to rotate according to the judgment result;

[0042] The information transmission device is built into the rotating mechanism, and is used to receive the judgment result generated by the central control platform according to the first judgment result and the second judgment result, obtain the target position information, and push the target position information to the central control platform.

[0043] Optionally, the central control platform includes an instruction analysis module and a big data risk assessment module;

[0044] The instruction analysis module is used to obtain the inspection instruction, analyze the inspection instruction to generate an execution action instruction, and send the execution action to the loading and unloading robot;

[0045] The big data risk assessment module is used to generate risk attribute quantified parameters of target goods according to customs declaration information, and display the risk attribute quantified parameters on the central control platform.

[0046] A second aspect of the present invention provides a verification processing method, wherein the verification processing method specifically includes:

[0047] Obtaining an inspection instruction, generating an execution action instruction according to the inspection instruction, the loading and unloading robot grabs the target goods according to the execution action instruction, and places the target goods on the conveying device, while the central control platform generates risk attribute quantification parameters of the target goods according to the customs declaration information obtained in advance, and displays the risk attribute quantification parameters on the central control platform;

[0048] The conveying device transmits the target goods to the inspection device, the inspection device performs size measurement, quantity counting and weight measurement on the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result, obtains a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device;

[0049] The mark recognition device is connected to the end of the inspection device, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device;

[0050] The separation device receives a judgment result generated by the central control platform according to the first judgment result and the second judgment result, transmits the target goods to a target location according to the judgment result, and pushes the target location information to the central control platform;

[0051] The central control platform generates risk attribute quantification parameters of the target goods according to the customs declaration information, and displays the risk attribute quantification parameters on the central control platform.

[0052] Optionally, the inspection device performs size measurement, quantity counting and weight measurement on the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result to obtain a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device, specifically including:

[0053] The conveying module transfers the target goods to the weighing module;

[0054] The weighing module measures the weight of the target cargo to obtain weighing information, and transmits the weighing information to the control cabinet;

[0055] The 3D camera measures the size and counts the quantity of the target goods to obtain size information and quantity information, and transmits the size information and quantity information to the control cabinet;

[0056] The control cabinet obtains the measurement result according to the weighing information, the quantity information and the size information, compares the parameter information with the measurement result to obtain the inspection result, determines whether the target cargo is abnormal according to the inspection result, obtains a first judgment result, determines whether to alarm according to the first judgment result, uploads the first judgment result to the central control platform, and transports the target cargo to the mark recognition device.

[0057] Optionally, the central control platform generates risk attribute quantification parameters of the target goods according to the pre-acquired customs declaration information, and displays the risk attribute quantification parameters on the central control platform, specifically including:

[0058] Inputting the customs declaration information into the trained classification model;

[0059] Generate risk attribute quantified parameters of target goods through the classification model, and display the risk attribute quantified parameters on the central control platform;

[0060] Wherein, the classification model is trained based on historical inspection data.

[0061] As can be seen from the above, in the system of the present invention, the integrated intelligent inspection system for box-type cargo includes a central control platform, a loading and unloading robot, a conveying device, an inspection device, a mark recognition device and a separation device; the central control platform is connected to the loading and unloading robot, the conveying device, the inspection device, the mark recognition device and the separation device through a network, and is used to obtain inspection instructions, parse the inspection instructions to generate execution action instructions, and send the execution actions to the loading and unloading robot, and at the same time, generate risk attribute quantification parameters of the target cargo according to the customs declaration information, and display the risk attribute quantification parameters on the central control platform; the loading and unloading robot is used to receive execution action instructions, grab the target cargo according to the execution action instructions, and place the target cargo on the conveying device; the conveying device is used to transfer the target cargo to the inspection device; the inspection device is connected to the end of the conveying device, and is used to measure the size, quantity and weight of the target cargo. , obtain a measurement result, compare the preset parameter information with the measurement result to obtain an inspection result, judge whether the target goods are abnormal according to the inspection result, obtain a first judgment result, upload the first judgment result to the central control platform, and transport the target goods to the mark recognition device; the mark recognition device is connected to the end of the inspection device, used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, judge whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device; the separation device is connected to the end of the mark recognition device, used to receive the judgment result generated by the central control platform according to the first judgment result and the second judgment result, transport the target goods to the target location according to the judgment result, and push the target location information to the central control platform.

[0062] Compared with the prior art, the current robots or automated equipment for logistics loading and unloading of customs cannot realize the processes of identifying, counting the quantity and identifying abnormal situations of goods, thus leading to the problem of long customs clearance time for goods. In the integrated intelligent inspection system for box-type goods proposed by the present invention, the inspection instructions obtained are converted into execution instructions through the central control platform, and the loading and unloading robots automatically move to the corresponding positions according to the obtained instructions to grab the target goods and place them on the corresponding conveying device. The conveying device transports the goods to the inspection device to determine whether the size and weight of the goods exceed the preset requirements, and once again determines whether there are any problems with the goods through the mark recognition device, and finally the abnormal and non-abnormal goods are transported separately through the separation device. Thereby, the goods are transported to the target location and the target location information is pushed to the central control platform. In addition, the type of target goods is judged according to the mark information and also pushed to the central control platform, so that the processes of cold chain grabbing, loading and unloading, transmission, inspection, dimension measurement, weighing, separation and information display can be automatically realized through the integrated system of intelligent inspection of box-type goods, which not only reduces the workload of the corresponding staff, but also this application processes the customs declaration information through the big data risk assessment module to generate the corresponding risk attribute quantitative parameters and display them to the user, which can provide a basis for users to carry out targeted risk control, reduce the inspection rate, save the inspection resource rate, greatly increase the operation efficiency, and realize the whole process intelligent inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0064] Figure 1 It is a schematic diagram of the principle of an integrated system for intelligent inspection of boxed cargo provided by an embodiment of the present invention;

[0065] Figure 2 It is a structural schematic diagram of a loading and unloading robot in a box-type cargo intelligent inspection integrated system provided by an embodiment of the present invention;

[0066] Figure 3 is a schematic diagram of the degrees of freedom of the gripping end provided by an embodiment of the present invention;

[0067] Figure 4 Schematic diagram of the mechanical principle of the grabbing end provided by an embodiment of the present invention;

[0068] Figure 5It is a schematic diagram of a target detection model in a visual perception module in an integrated system for intelligent inspection of box-type cargo provided by an embodiment of the present invention;

[0069] Figure 6 It is a schematic diagram of a conveying device in the integrated system for intelligent inspection of box-type cargo provided by an embodiment of the present invention;

[0070] Figure 7 Schematic diagram of a roller conveyor belt in a conveying device in an integrated system for intelligent inspection of box-type cargo provided by an embodiment of the present invention;

[0071] Figure 8 Schematic diagram of an inspection device in an integrated intelligent inspection system for box-type cargo provided by an embodiment of the present invention;

[0072] Fig. 9 Schematic diagram of a mark recognition device in an integrated system for intelligent inspection of boxed cargo provided by an embodiment of the present invention;

[0073] Fig.10 It is a schematic diagram of a mark recognition device combined with a separation device in an integrated system for intelligent inspection of box-type cargo provided by an embodiment of the present invention;

[0074] Fig.11 It is a workflow diagram of an inspection and processing method provided by an embodiment of the present invention.

[0075] Figure numerals: 1, central control platform; 2, loading and unloading robot; 3, conveying device; 4, inspection device; 5, mark recognition device; 6, separation device; 21, chassis; 22, hybrid robot arm; 23, grab end; 24, visual perception module; 25, vehicle body control box; 26, robot conveyor belt; 261, primary conveyor belt; 262, secondary telescopic conveyor belt; 231, bottom support; 232, vacuum suction cup; 233, conveyor belt; 31, machine Frame; 32, back roller conveyor belt; 33, lifting module; 34, roller conveyor belt; 41, conveying module; 42, weighing module; 43, 3D camera; 44, control cabinet; 45, sound and light alarm module; 51, mark recognition frame; 52, industrial camera; 53, motion mechanism; 54, rotating platform; 55, mark recognition conveying module; 56, second baffle; 57, first baffle; 58, pressure plate; 61, roller conveyor belt; 62, rotating mechanism. DETAILED DESCRIPTION

[0076] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0077] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0078] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0079] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0080] As used in this specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to being classified into," depending on the context. Similarly, the phrase "if it is determined" or "if classified into [described condition or event]" may be interpreted as meaning "upon determination" or "in response to determining" or "upon classification into [described condition or event]" or "in response to being classified into [described condition or event]," depending on the context.

[0081] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. 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.

[0082] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0083] The integrated intelligent inspection system for boxed cargo described in the preferred embodiment of the present invention is as follows: Figure 1 As shown, the integrated intelligent inspection system for boxed cargo includes a central control platform 1, a loading and unloading robot 2, a conveying device 3, an inspection device 4, a mark recognition device 5 and a separation device 6;

[0084] The central control platform 1 is connected to the cargo handling robot 2, the conveying device 3, the inspection device 4, the mark recognition device 5 and the separation device 6 through a network, and is used to obtain inspection instructions, parse the inspection instructions to generate execution action instructions, and send the execution action to the cargo handling robot 2. At the same time, according to the customs declaration information obtained in advance, the risk attribute quantitative parameters of the target goods are generated, and the risk attribute quantitative parameters are displayed on the central control platform 1;

[0085] The loading and unloading robot 2 is used to receive an execution action instruction, grab the target goods according to the execution action instruction, and place the target goods on the conveying device 3;

[0086] The conveying device 3 is used to transport the target goods to the inspection device 4;

[0087] The inspection device 4 is connected to the end of the conveying device 3, and is used to measure the size, quantity and weight of the target goods, obtain the measurement results, compare the preset parameter information with the measurement results, obtain the inspection results, judge whether the target goods are abnormal according to the inspection results, obtain the first judgment result, upload the first judgment result to the central control platform 1, and convey the target goods to the mark recognition device 5;

[0088] The mark recognition device 5 is connected to the end of the inspection device 4, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform 1, and transport the target goods to the separation device 6;

[0089] The separation device 6 is connected to the end of the mark recognition device 5, and is used to receive the judgment result generated by the central control platform 1 based on the first judgment result and the second judgment result, transport the target goods to the target location according to the judgment result, and push the target location information to the central control platform 1.

[0090] The central control platform 1 converts the obtained inspection instruction into an execution action instruction, and controls the loading and unloading robot 2 to place the target goods on the conveying device 3 by executing the control instruction. After the conveying device 3 conveys the goods to the inspection device 4, the inspection device 4 measures the size, counts the quantity and measures the weight of the target goods to obtain the measurement result, compares the pre-set parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result, obtains the first judgment result, uploads the first judgment result to the central control platform 1, and conveys the target goods to the mark recognition device 5; the mark recognition device 5 compares the information of the goods again to obtain the second judgment result, uploads the second judgment result and the mark information to the central control platform 1, and conveys the target goods to the separation device 6; and when the central control platform 1 obtains the inspection instruction, it generates the risk attribute quantification parameter of the target goods according to the pre-acquired customs declaration information, and displays the risk attribute quantification parameter on the central control platform 1.

[0091] Among them, in the central control platform 1, the staff can fill in the inspection instructions (such as which platform to inspect) and customs declaration information (such as origin, product name, specifications) in the upper computer interface of the central control platform 1 according to the format. The central control platform 1 is connected to each device, namely the loading and unloading robot 2, the conveying device 3, the inspection device 4, the mark recognition device 5 and the separation device 6 through the network. In one embodiment of the present application, the connection is through EtherCat communication. After the central control platform 1 receives the specific inspection action, it will send a control instruction to the loading and unloading robot 2, so that the loading and unloading robot 2 moves to the inspection platform, connects and starts each device, takes pictures to calculate the grabbing point, the robot arm moves, unloads, circulates, inspects, marks recognize, separates and the customs staff conducts on-site verification. The central control platform 1 includes an industrial computer and a developed software system. The central control platform 1 can control and adjust the loading and unloading robot 2, the conveying device 3, the inspection device 4, the mark recognition device and the separation device 6. It will monitor the status of each device in real time and alarm and handle abnormal conditions in each device. At the same time, the central control system also provides an operation and human-computer interaction interface, so that users can easily view the system status, operate equipment, set parameters, etc. This allows the operator to understand the system operation more intuitively and perform necessary operations and adjustments.

[0092] Specifically, Figure 2 As shown, the loading and unloading robot 2 includes a chassis 21, a vehicle body control box 25, a hybrid robot arm 22, a gripping end 23, a visual perception module 24 and a control module;

[0093] The chassis 21 is located at the bottom of the loading and unloading robot 2 and is used to drive the loading and unloading robot 2 to move;

[0094] The vehicle body control box 25 is fixedly disposed on the chassis 21 and is used to load the control module and support the visual perception module 24;

[0095] The hybrid robot arm 22 is disposed on the chassis 21 and is used to drive the grasping end 23 to move in space;

[0096] The grabbing end 23 is disposed at the end of the hybrid robotic arm 22 and is used to absorb and transfer the target cargo;

[0097] The visual perception module 24 is disposed on the hybrid robotic arm 22 and is used to obtain the grasping posture of the target cargo;

[0098] The control module is disposed on the vehicle body control box 25, and is used to control the chassis 21, the hybrid mechanical arm 22, and the grabbing end 23 to absorb the target cargo and then transfer it to the conveying device 3;

[0099] The control module is electrically connected to the chassis 21 , the visual perception module 24 , the hybrid robot arm 22 and the grasping end 23 , respectively.

[0100] In one embodiment of the present invention, the loading and unloading robot 2 may be a fully automatic loading and unloading robot. The fully automatic loading and unloading robot may realize all functions of the loading and unloading robot 2. Figure 2 As shown, the loading and unloading robot 2 also includes a robot conveyor belt 26, wherein the robot conveyor belt is composed of a primary conveyor belt 261 and a secondary telescopic conveyor belt 262, the primary conveyor belt 261 is arranged on the hybrid robotic arm 22, and the secondary telescopic conveyor belt 262 is arranged between the hybrid robotic arm 22 and the vehicle body control box 25, and is used for connecting and conveying target goods; further, the electrical connection between the operating arm module, the visual perception module and the central control platform is realized in the vehicle body control box.

[0101] Further, such as Figure 4 As shown, the grabbing end 23 includes a base 231, a vacuum suction cup 232 and a conveyor belt 233;

[0102] The base 231 is used to support the vacuum suction cup 232 and the conveyor belt 233;

[0103] The vacuum suction cup 232 is slidably disposed on the base 231 through a slider, and is used to absorb the target goods onto the conveyor belt 233;

[0104] The conveyor belt 233 covers the base 231 and is used to transport the target goods to the conveying device 3 .

[0105] Specifically, through the method of combining side adsorption with bottom support, the vacuum suction cup 232 can move forward and backward to drag the goods to the bottom support 231, and the weight of the goods is provided by the support force of the bottom support 231, wherein the bottom support 231 is the base of the conveyor belt 233 and can play a supporting role. However, when grabbing the bottom layer of goods, the bottom support 231 at a certain height will interfere with the side of the cargo box, so the vertical freedom of movement of the vacuum suction cup 232 is increased so that it can lift the goods for a certain distance first. In addition, through the conveyor belt 233 with the bottom support 231, the friction between the goods and the conveyor belt becomes the main force, which can grab the target cargo box more efficiently and reliably. Therefore, it can be concluded from the above that the end grasping mechanism has at least three degrees of freedom, namely the vertical and horizontal freedom of movement of the vacuum suction cup 232 and the rotational movement of the conveyor belt. The specific degrees of freedom are as follows: Figure 3 As shown in . Furthermore, in the gripping end 23, the vertical movement freedom of the vacuum suction cup 232 adopts a parallel four-bar linkage combined with a cam guide mechanism for setting the lifting trajectory, and the horizontal movement freedom of the vacuum suction cup 232 adopts a linear guide mechanism, and the movement of the two is completed by the same motor driving the linear guide slider, that is, during the horizontal movement of the gripping end 23 mechanism, the parallel four-bar linkage will also be passively deformed along with the cam guide mechanism, thereby realizing the vertical lifting movement of the vacuum suction cup 232; the rotational movement of the conveyor belt is driven by an electric roller.

[0106] Furthermore, the visual perception module 24 includes a laser radar depth camera and a neural network, and the laser radar obtains the depth information of the object to be measured and obtains the grasping posture of the target cargo. Among them, the laser radar depth camera is used to take a photo of the cargo box placed in the container, and the neural network of the Mask·R-CNN architecture, that is, the target detection model, is used to identify and pixel-level segment the cargo box to obtain a mask result, and the image pixel area is extracted from the mask result, and the corresponding mask area is converted into a three-dimensional point cloud. The grabbing plane of the packaging box is extracted by a random sampling consensus algorithm, and the local points belonging to the plane are used as the final valid point set. The algorithm for calculating the minimum area enclosing rectangle of the point set is used to obtain the rectangular boundary of the top surface of the cargo grab; the depth information returned by the laser radar depth camera scan is obtained, and the spatial position of the cargo grabbing surface is calculated by the RANSAC algorithm, that is, the grasping posture of the target cargo. The control module controls the chassis 21, the hybrid robot arm 22 and the grabbing end 23 to absorb the target cargo and then transfer it to the conveying device 3 through the spatial position of the target cargo and the rectangular boundary of the cargo grabbing top surface.

[0107] Specifically, for the detection and positioning of densely stacked box cargo, the object detection model is used to extract the grasping pose. Specifically: first, the detection and segmentation branches are intertwined for joint multi-stage processing, and the network framework that uses the semantic segmentation branch to provide spatial context information is more suitable for mask prediction of dense targets; therefore, Figure 5 As shown, the instance segmentation network based on the hybrid task cascade (HTC) is improved Figure 5 The target detection model shown in the figure specifically takes the RGB image and the corresponding depth information as input. In view of the noise and depth missing problems of the RGBD depth map, the median filter is used to remove the noise of the original depth map and the weighted joint bilateral filter is used to fill the hole area, and then a simulated interaction layer is added to produce an interactive image. Then, the feature map extracted by the feature extraction module of Resnet101 as the backbone network is input into the feature pyramid FPN, and a feature pyramid structure is added after the feature pyramid to form a double feature pyramid structure to enhance the spatial information of the feature maps of each level. After taking out the corresponding feature area from each layer, a feature fusion layer is added to fuse the feature maps of all levels and input them into the subsequent target mask prediction network; at the same time, semantic segmentation is introduced into the instance segmentation framework, and the semantic segmentation results and semantic segmentation features are predicted by fusion and convolution of each feature map of FPN, and the features are fused with the original mask to obtain better spatial information. In addition, a MaskIOU is added at the end of the network model. The head module takes the instance features and the predicted mask as input, uses the maximum pooling layer to make the predicted mask have the same spatial size as the ROI feature, and then regresses the IOU between the predicted mask and the ground truth mask through 3 convolutional layers and 3 fully connected layers to obtain a more reliable motion mask score, thereby improving the network prediction accuracy.

[0108] For the grasping surface, the region is extracted based on the target detection model result; the mask result of the corresponding packaging box is obtained according to the prediction of the target detection model, and the corresponding image pixel area is taken out; the corresponding mask area is converted into a 3D point cloud and processed using the depth map to 3D point cloud algorithm. Specifically, the correspondence between a spatial point [x, y, z] and its pixel coordinates [u, v, d] in the depth map to 3D point cloud algorithm is shown in the following formula:

[0109] ;

[0110] ;

[0111] ;

[0112] in, , , refers to the focal length of the camera on the x and y axes, , , refers to the aperture center of the camera, and s refers to the scaling factor of the depth map. The process of deriving (u,v,d) from (x,y,z) can be expressed as the following formula:

[0113] .

[0114] Then, the random sampling consensus algorithm (RANSAC algorithm) is used to extract the grasping plane of the packaging box, and the points inside the plane are taken as the final valid point set. The algorithm for calculating the minimum area enclosing rectangle of the point set is used to estimate the rectangular boundary of the top surface of the packaging box. The specific steps of the RANSAC algorithm include:

[0115] Step 1: Randomly select three points from the point cloud;

[0116] Step 2: These three points form a plane;

[0117] Step 3: Calculate the distance from all other points to the plane. If the distance is less than the threshold T, the points are considered to be in the same plane.

[0118] Step 4: If there are more than n points on the same plane, save the plane, mark all points on the plane as matched, and repeat steps 1 to 4.

[0119] Step 5: Iterate steps 1 to 4 above until the plane is smaller than n points or no three unmarked points are found.

[0120] In addition, the loading and unloading robot 2 also includes a control system and a sensor module. The control system is responsible for coordinating the chassis 21, the hybrid robot arm 22, the gripping end 23, the vision module, etc., to execute the control instructions of the operation process. The sensor module is set on the robot arm and is divided into three categories, namely, photoelectric sensors, which are used to detect the cargo boxes passing on the conveyor belt; ranging sensors, including laser ranging and ultrasonic ranging, which provide the robot arm with the distance in the forward direction, left and right direction, and up and down direction; safety sensors, laser scanners prevent the robot arm from colliding with people and walls, and limit sensors ensure that the robot arm operates in a safe working space to prevent accidents.

[0121] like Figure 6 and Figure 7 As shown, the conveying device 3 includes a frame 31, a back-rolling conveyor belt 32, a lifting module 33 and a roller conveyor belt 34;

[0122] The frame is used to fix the back roller conveyor belt 32, the lifting module 33 and the roller conveyor belt 34;

[0123] The back roller conveyor belt 32 and the lifting module 33 are slidably arranged on the frame 31, the roller conveyor belt 34 is detachably arranged on the frame 31, the roller conveyor belt 34 and the lifting module 33 are arranged on the same side of the frame 31, wherein the back roller conveyor belt 32 is arranged on the opposite side of the roller conveyor belt 34 and the lifting module 33;

[0124] The back-rolling conveyor belt 32 is used to place the target goods;

[0125] The lifting module 33 is used to adjust the height of the back roller conveyor belt 32 to the height of the roller conveyor belt 34;

[0126] The roller conveyor belt 34 is used to convey the target goods to the inspection device 4 .

[0127] Specifically, after the target goods are grabbed by the loading and unloading robot 2 and placed on the conveying device 3, the conveying device 3 transports the goods from the back roller conveyor belt 32 to the roller conveyor belt 34 through the lifting module 33, and the goods are transferred to the inspection device 4 by the roller conveyor belt 34.

[0128] Further, such as Figure 6 and Figure 7 As shown, the frame 31 further includes a connecting member 35 for detachably mounting the roller conveyor belt 34 on the frame 31 .

[0129] Specifically, Figure 8 As shown, the inspection device 4 includes a conveying module 41, a weighing module 42, a 3D camera 43, a control cabinet 44 and an audible and visual alarm module 45;

[0130] The conveying module is used to transfer the target goods to the weighing module 42;

[0131] The weighing module 42 is disposed below the conveying module 41 and is used to measure the weight of the target goods, obtain weighing information, and transmit the weighing information to the control cabinet 44;

[0132] The 3D camera is fixedly mounted on the control cabinet 44, and is used to measure the size and count the quantity of the target goods, obtain size information and quantity information, and transmit the size information and quantity information to the control cabinet 44;

[0133] The control cabinet 44 is disposed beside the conveying module 41, and is used to obtain the measurement result according to the weighing information, quantity information and size information, compare the parameter information with the measurement result to obtain the inspection result, judge whether the target cargo is abnormal according to the inspection result, obtain a first judgment result, and upload the first judgment result to the central control platform 1;

[0134] The sound and light alarm module 45 is arranged on the control cabinet and is used to determine whether to give an alarm according to the inspection result;

[0135] The control cabinet 44 is electrically connected to the conveying module 41 , the weighing module 42 , the 3D camera 43 and the sound and light alarm module 45 , respectively.

[0136] Specifically, the conveying module 41 drives the conveyor belt to run through the power roller, so that the target goods can pass on the conveyor belt. The weighing module 42 relies on the weighing sensor below to weigh the target goods passing through the conveying module 41. After the weighing sensor obtains the analog signal, it performs A / D conversion, and the measurement result outputs the corresponding weighing information through the bus serial mode; at the same time, the 3D camera 43 obtains the depth information and RGB image of the measured object, so as to obtain the three-dimensional size of the cargo box and obtain the size information. An industrial computer is built in the control cabinet 44, which is used to count and record the size, weight, quantity and other parameter information of all the inspection equipment for traceability and inspection and comparison, and upload these parameter information to the central control platform 1. At the same time, parameter information is also preset, which is the weight, quantity and size information of the batch of goods that pass the customs, that is, the standard range of the goods. By comparing the parameter information with the measurement result, the inspection result is obtained. Once the inspected goods exceed the standard range, the sound and light alarm module 45 will quickly alarm, which can effectively intercept the abnormality. At the same time, whether the target goods are abnormal is determined according to the inspection result to obtain a first judgment result, the first judgment result is uploaded to the central control platform 1, and the target goods are transported to the mark recognition device 5.

[0137] Further, the sound and light alarm module 45 receives the inspection result; when the inspection result is the first inspection result, the sound and light alarm module 45 issues an alarm; when the inspection result is the second inspection result, the sound and light alarm module 45 does not issue an alarm. Wherein, when the inspection result shows that the measurement result does not meet the parameter information requirements, the inspection result is the first inspection result; when the inspection result shows that the measurement result meets the parameter information requirements, the inspection result is the second inspection result. When the inspection result is the first inspection result, the first judgment result is abnormal; when the inspection result is the second inspection result, the first judgment result is normal.

[0138] Specifically, Fig. 9 As shown, the mark recognition device 5 includes a mark recognition frame 51, an industrial camera 52, a motion mechanism 53, a rotating platform 54 and a mark recognition conveying module 55;

[0139] The mark recognition frame 51 is used to fix the rotating platform 54 and the motion mechanism 53;

[0140] The motion mechanism 53 is disposed on the mark recognition frame 51 and is used to rotate the target goods;

[0141] The industrial camera 52 is used to take an initial photo of the target goods to obtain an initial position photo, control the motion mechanism 53 to move to the mark position according to the initial position photo, and take a photo of the mark of the target goods to obtain a mark position photo, extract the mark information of the target goods according to the mark position photo, compare the pre-acquired customs declaration information with the mark information, judge whether the target goods are abnormal, obtain a second judgment result, and upload the second judgment result and the mark information to the central control platform 1;

[0142] The movement mechanism 53 is disposed on the mark recognition frame 51 and is used to move according to the initial position photo;

[0143] The mark recognition and conveying module 55 is disposed on the mark recognition frame 51 and is used to convey the target goods to the separation device 6 .

[0144] Specifically, the mark recognition device 5 includes an industrial camera 52, a 3DOF motion mechanism 53, a rotating platform 54, a mark recognition conveying module 55, a first baffle 57, a second baffle 56 and a pressing plate 58. The target goods move to the rotating platform 54 through the conveying module 41 in the inspection device 4 until the first baffle 57 is in close contact with one side of the target goods, and the pressing plate 58 presses the other side of the cargo box to be flush with the second baffle 56, and then the rotating platform 54 rotates 90° counterclockwise to make the mark label surface of the target goods face the industrial camera 52. The industrial camera 52 is installed on the motion mechanism 53, and is automatically aligned by taking pictures. The alignment steps specifically include: the cargo information obtained through the central control platform 1 can be used to know the width D of the cargo, and a short focal length, large depth of field camera is used. Through the experience value setting, the target cargo (the entire cargo box) is fully exposed to the camera field of view, and the first photo is taken. The cargo box is imaged by the character recognition module camera, while ensuring the clarity and quality of the image; combined with target detection and image processing technology, the pixel coordinate values ​​of the target cargo can be obtained, which are the width pixel coordinate value and height pixel coordinate value of the cargo box (i.e., expressed as box_width, box_height), and the pixel coordinate value of the mark label, wherein the pixel coordinate value of the mark label includes the leftmost pixel point of the label, the highest pixel point of the label, the pixel coordinate value of the label width, and the pixel coordinate value of the label height (i.e., expressed as label_left, abel_to p, label_width, label_height); according to unit=D / box_width, get the physical size corresponding to each pixel; according to unit*label_width, unit*label_height, get the length and width of the label (the label is 2D, the length of the label is width, and the width is height); then find the center coordinates of the label, according to center_x=label_left+0.5*label_width, center_y=label_top+0.5*label_hight, get the center coordinates (x, y); finally, solve the difference between the center coordinates of the label and the center coordinates of the photo, that is, the center coordinates of the photo are (pic_width / 2, pic_height / 2), delta_x=(center_x- pic_width / 2)*unit, delta_y=(center_x-pic_height / 2)*unit, where pic represents the captured image, and obtains the difference between the coordinates of the center point of the label and the coordinates of the center point of the photo; after obtaining the x coordinate delta_x and y coordinate delta_y of the center point of the label and the center point of the photo, the controller controls the conveying screws in the x, y and z directions of the motion mechanism 53 to move the camera to the position of the mark so that the camera is aligned with the mark label.

[0145] The camera takes a second photo of the mark, and then performs OCR recognition on the mark information. Specifically, the photograph is uploaded to a third-party cloud, and after OCR recognition in the cloud, the key information (origin, product name, specifications) is classified and extracted, and structured information is output. The customs declaration information, such as origin, product name, quantity, specifications (length, width, height and weight, etc.), is compared one by one with the structured information (i.e., the information identified by the mark information). For abnormal goods, the origin on the mark label is inconsistent with the customs declaration information, etc. In one embodiment of the present application, an alarm is set in the mark recognition device 5. An alarm is sounded through the alarm, and the separation is carried out in the next separation device 6, and uploaded to the central control platform 1 for recording and statistics for analysis and tracing.

[0146] Furthermore, the pre-acquired customs declaration information is compared with the mark information, that is, the structured information is compared with the customs declaration information to determine whether the target goods are abnormal. When the structured information is different from the customs declaration information, the second judgment result is abnormal. When the structured information is the same as the customs declaration information, the second judgment result is normal. The second judgment result and the mark information are uploaded to the central control platform 1.

[0147] Specifically, Fig.10 As shown, the separation device 6 includes a roller conveyor belt 61, a rotating mechanism 62 and an information transmission device;

[0148] The roller conveyor belt 61 is connected to the rotating mechanism 62 and is used to convey the target goods to the target location;

[0149] The rotating mechanism 62 is used to rotate according to the judgment result;

[0150] The information transmission device is built into the rotating mechanism 62 , and is used to receive the judgment result generated by the central control platform 1 according to the first judgment result and the second judgment result, obtain the target position information, and push the target position information to the central control platform 1 .

[0151] Specifically, Fig.10 As shown, a roller conveyor is provided in the mark recognition device 5 and the separation device 6 to convey the target goods. The separation device 6 is composed of a roller conveyor 61 and a rotating mechanism 62. The rotating mechanism 62 is controlled by a motor. For normal goods, the rotating mechanism 62 does not rotate, and the target goods flow normally to the next link and are transmitted to the first target position. For abnormal goods, the rotating mechanism 62 rotates 90°, and the target goods are transported to the abnormal goods processing area, that is, the second target position.

[0152] Among them, it should be noted that the rotating mechanism 62 has a built-in information transmission device, which is used to receive the judgment result generated by the central control platform 1 according to the first judgment result and the second judgment result, and rotate according to the judgment result. Specifically, when the first judgment result and the second judgment result are both normal goods, the rotating mechanism 62 does not rotate, and the target goods flow normally to the next link and are transmitted to the first target position; otherwise, the rotating mechanism 62 rotates 90°, and the target goods are transported to the area for abnormal goods processing, that is, the second target position. In the separation device 6, the abnormal goods screened by the inspection device 4 and the mark recognition device are separated from the normal goods by the separation device 6 and divided into two piles. Further, when the user obtains the target goods placed in the second target position on the central control platform 1, a manual re-inspection can be performed accordingly to ensure the integrity of the abnormal situation identification. The inspector can further verify the abnormal situation of the goods according to the information of the central control platform 1.

[0153] Furthermore, the central control platform includes an instruction analysis module and a big data risk assessment module;

[0154] The instruction analysis module is used to obtain the inspection instruction, analyze the inspection instruction to generate an execution action instruction, and send the execution action to the loading and unloading robot;

[0155] The big data risk assessment module is used to generate risk attribute quantified parameters of target goods according to customs declaration information, and display the risk attribute quantified parameters on the central control platform.

[0156] Further, in one embodiment of the present application, the instruction analysis module and the big data risk assessment module are placed on the central control machine of the central control platform. The customs declaration information and the mark information specifically include the origin of the goods, the type of goods and other information, and the risk attribute quantification parameters include risk category and risk degree, the risk category includes abnormal goods name, abnormal goods quantity, abnormal goods specification and abnormal goods origin, etc., and the risk degree corresponds to the risk category, and the weight of the goods with such risk. In one embodiment of the present application, the basa fish in the first target area is processed as the object, so as to carry out targeted control according to each batch of goods. When the abnormal risk index of the quantity of basa fish in the first target area is 6 obtained through the decision tree C5.0 model, the risk attribute quantification parameter of the basa fish in the first target area is the abnormal risk of the quantity of goods, and the corresponding personnel can carry out targeted inspection and control of the basa fish from the first target area, focusing on counting the quantity.

[0157] Among them, the customs declaration information is transmitted to the classification model, and the risk attribute quantitative parameters of the target goods are output; and the risk attribute quantitative parameters are displayed on the central control platform. In addition, the training process of the classification model specifically includes: obtaining a preliminary training data set, and performing feature engineering on the preliminary training data set to obtain risk attribute quantitative parameters, then labeling each data sample, generating a risk label, obtaining an available training data set and an available test data set, and training the classification model through the available training data set; when the classification model reaches the preset training conditions, the training is stopped, and the classification model obtained from the last training is used as the completed classification model and output. Among them, the classification model selects the decision tree C5.0 model, and the preliminary model is established using Python's scikit-learn package. The parameter selection uses the default parameters as a whole, and the tree depth is selected as 16.

[0158] Furthermore, during the training process, obtaining a preliminary training data set includes obtaining all import and export goods data in a database to form a preliminary training data set, including specific information such as the origin of the goods, the type of goods, the quantity of goods, the specifications of the goods, the quantity of goods, the weight of goods, and the size of goods.

[0159] Afterwards, a training data set is obtained through feature engineering. That is, feature engineering is performed on the data of the preliminary training data set, and the method of text feature extraction and natural language processing technology are used to identify risk-related information. First, entity recognition technology is used to extract risk-related entities from the text, such as cargo type, origin, risk index, etc. This helps to determine the specific objects discussed in the text, and helps to analyze the risk information related to these entities and extract risk-related features. At the same time, text mining technology is used to process the unstructured information related to the cargo in the data statistics. Among them, the steps of obtaining unstructured information include: data collection and cleaning, collecting unstructured text data containing cargo information, cleaning data to remove noise, such as HTML tags, special characters or other irrelevant information; performing word segmentation, dividing the text into words or phrases; then performing stem extraction and lemmatization, stem extraction truncates words to their basic form, and lemmatization restores words to their original form; further, performing entity recognition, identifying entities in the text, such as product name, origin, weight, etc., which is very important for understanding the specific cargo information in the text; then performing keyword extraction, capturing the core information of the text by extracting keywords or phrases in the text. The TF-IDF (term frequency-inverse document frequency) method can be used to measure the importance of words in the text; the use of topic modeling techniques (such as Latent Dirichlet Allocation, LDA) to discover the topic structure in the text helps to understand the potential topics and relevance of the cargo description. Using a supervised learning classification algorithm, the text is divided into different categories or groups. This helps to organize and understand large amounts of text data. Finally, according to specific needs, unstructured information related to the goods is extracted from the processed text, such as the origin, size, weight, etc. of the goods.

[0160] Then, the training data set is labeled with data labels to obtain an available training data set. That is, according to historical data, the corresponding risk attribute quantification parameters are labeled for the samples of each training data set, that is, multi-dimensional risk labels are labeled for each sample, and the label content includes the name of the goods, the quantity of goods, the specifications of goods, the origin of goods, the risk category, etc., which will be used as the training target of the classification model. Among them, in one embodiment of the present application, for durian, the format of the label can be expressed as: [Name: Durian, Quantity: 1200 boxes, Specifications (length x width x height): 300x150x400, Weight: 15kg, Origin: Target area, Risk attribute quantification parameter: Abnormal weight risk]. After labeling, an available training data set is obtained, and the decision tree C5.0 model is trained by the available training data set. Among them, during the training process, the present invention mines and extracts 80% of the data as a training set, and 20% of the data as a test set. The training set is used to train the model, and the test set is used to evaluate the performance of the model. After each training is completed, an evaluation is performed, and the performance of the model is evaluated using the test set. Determine whether to continue training based on the evaluation results. When the decision tree C5.0 model reaches the preset training conditions, stop training, and use the decision tree C5.0 model obtained from the last training as the trained decision tree C5.0 model and output it; the preset training condition is that the performance of the decision tree C5.0 model reaches a preset value. Preferably, in one embodiment of the present application, stop training when the accuracy rate reaches 98% or more.

[0161] Furthermore, in this application, for the trained decision tree C5.0 model, historical data can be obtained regularly according to user settings to construct a training data set, so as to train the decision tree C5.0 model again, so that the decision tree C5.0 model can continue to learn. The model is trained using historical data, and the improvement of model performance mainly relies on the continuous collection of high-quality incremental data and the existing data to be trained and updated at a certain frequency. The big data risk assessment module obtains risk assessment results after obtaining the customs declaration information, and then conducts risk control.

[0162] The current customs logistics loading and unloading robots or automated equipment are unable to realize the processes of identifying, counting and distinguishing abnormal situations of goods, which leads to the problem of long customs clearance time. In the integrated intelligent inspection system for box-type goods proposed by the present invention, the inspection instructions obtained are converted into execution instructions through the central control platform, and the loading and unloading robots automatically move to the corresponding position according to the obtained instructions to grab the target goods and place them on the corresponding conveying device. The conveying device transports the goods to the inspection device to determine whether the size and weight of the goods exceed the preset requirements, and once again determines whether there are any problems with the goods through the mark recognition device. Finally, the abnormal and non-abnormal goods are transported separately through the separation device, so as to separate the goods. It is transported to the target location and the target location information is pushed to the central control platform. In addition, the type of target goods can be determined based on the mark information and is also pushed to the central control platform, so that the cold chain grabbing, loading and unloading, transmission, inspection, dimension measurement, weighing, separation and information display processes can be realized automatically through the integrated intelligent inspection system for box-type cargo, which not only reduces the workload of the corresponding staff, but also processes the customs declaration information through the big data risk assessment module to generate the corresponding risk attribute quantitative parameters and display them to the user, which can provide a basis for users to carry out targeted risk control, reduce the inspection rate, save inspection resources, greatly increase the operating efficiency, and realize full-process intelligent inspection.

[0163] Further, based on Figure 1 The box-type cargo intelligent inspection integrated system shown in the figure, the preferred embodiment of the present invention proposes an inspection and processing method for the box-type cargo intelligent inspection integrated system, such as Fig.11 As shown, the inspection processing method includes the following steps:

[0164] Step S10: the central control platform obtains the inspection instruction, generates an execution action instruction according to the inspection instruction, the loading and unloading robot grabs the target goods according to the execution action instruction, and places the target goods on the conveying device, and at the same time, the central control platform generates risk attribute quantification parameters of the target goods according to the customs declaration information obtained in advance, and displays the risk attribute quantification parameters on the central control platform;

[0165] Step S20, the conveying device transmits the target goods to the inspection device, the inspection device performs size measurement, quantity counting and weight measurement on the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result, obtains a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device;

[0166] Step S30, the mark recognition device is connected to the terminal of the inspection device, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device;

[0167] Step S40, the separation device receives the judgment result generated by the central control platform according to the first judgment result and the second judgment result, transports the target goods to the target location according to the judgment result, and pushes the target location information to the central control platform.

[0168] Further, the inspection device performs size measurement, quantity counting and weight measurement on the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result to obtain a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device, specifically including:

[0169] The conveying module transfers the target goods to the weighing module;

[0170] The weighing module measures the weight of the target cargo to obtain weighing information, and transmits the weighing information to the control cabinet;

[0171] The 3D camera measures the size and counts the quantity of the target goods to obtain size information and quantity information, and transmits the size information and quantity information to the control cabinet;

[0172] The control cabinet obtains the measurement result according to the weighing information, the quantity information and the size information, compares the parameter information with the measurement result to obtain the inspection result, determines whether the target cargo is abnormal according to the inspection result, obtains a first judgment result, determines whether to alarm according to the first judgment result, uploads the first judgment result to the central control platform, and transports the target cargo to the mark recognition device.

[0173] The central control platform generates risk attribute quantification parameters of target goods according to the pre-acquired customs declaration information, and displays the risk attribute quantification parameters on the central control platform, specifically including:

[0174] Obtaining customs declaration information, and inputting the customs declaration information into the trained classification model;

[0175] Generate risk attribute quantified parameters of target goods through the classification model, and display the risk attribute quantified parameters on the central control platform;

[0176] Wherein, the classification model is trained based on historical inspection data.

[0177] In the present invention, the inspection instruction obtained is converted into an execution action instruction through the central control platform, and the loading and unloading robot automatically moves to the corresponding position according to the obtained instruction to grab the target goods and place them on the corresponding conveying device. The conveying device conveys the goods to the inspection device to determine whether the size and weight of the goods exceed the preset requirements, and once again determines whether there is a problem with the goods through the mark recognition device. Finally, the abnormal and non-abnormal goods are transported separately through the separation device, so that the goods are transported to the target position, and the target position information of the transport is pushed to the central control platform. In addition, when the type of target goods can be exceeded according to the mark information, it is also pushed to the central console, so that the process of cold chain grabbing, loading and unloading, transmission, inspection, size measurement, weighing, separation and information display can be realized automatically through the integrated intelligent inspection system for box-type goods, thereby not only reducing the workload of the corresponding staff, but also the application processes the customs declaration information through the big data risk assessment module to generate corresponding risk attribute quantitative parameters and display them to the user, so that the user can control the risk, which greatly increases the work efficiency and provides convenience for the user.

[0178] Furthermore, the present invention may change the design, for example, using a serial robot arm instead of a hybrid robot arm; or using a 2D lidar + reflector, or an RFID + mark point to realize the navigation and autonomous movement of the loading and unloading robot; or in the inspection equipment, a ranging sensor is installed on the top and both sides, the sensor on the top can measure the height and length of the cargo box, and the sensors on both sides can measure the width of the cargo box. In this way, the size information of the target cargo can also be detected.

[0179] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

[0180] It should be understood that the serial numbers of the steps in the above embodiments do not imply a sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0181] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the above-mentioned device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0182] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0183] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0184] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features therein may be replaced by equivalents. However, these modifications or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. An integrated intelligent inspection system for boxed cargo, characterized in that: The integrated intelligent inspection system for boxed cargo includes a central control platform, a loading and unloading robot, a conveying device, an inspection device, a mark recognition device and a separation device; The central control platform is connected to the loading and unloading robot, the conveying device, the inspection device, the mark recognition device and the separation device through a network, and is used to obtain inspection instructions, parse the inspection instructions to generate execution action instructions, and send the execution actions to the loading and unloading robot. At the same time, according to the customs declaration information obtained in advance, the risk attribute quantitative parameters of the target goods are generated, and the risk attribute quantitative parameters are displayed on the central control platform; The loading and unloading robot is used to receive an execution action instruction, grab the target goods according to the execution action instruction, and place the target goods on the conveying device; The conveying device is used to transport the target goods to the inspection device; The inspection device is connected to the end of the conveying device, and is used to measure the size, quantity and weight of the target goods, obtain a measurement result, compare the preset parameter information with the measurement result, obtain an inspection result, judge whether the target goods are abnormal according to the inspection result, obtain a first judgment result, upload the first judgment result to the central control platform, and convey the target goods to the mark recognition device; The mark recognition device is connected to the end of the inspection device, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device; The separation device is connected to the end of the mark recognition device, and is used to receive the judgment result generated by the central control platform according to the first judgment result and the second judgment result, and to transfer the target goods to the target location according to the judgment result, and to push the target location information to the central control platform; The central control platform includes an instruction analysis module and a big data risk assessment module; The instruction analysis module is used to obtain the inspection instruction, analyze the inspection instruction to generate an execution action instruction, and send the execution action to the loading and unloading robot; The big data risk assessment module is used to generate risk attribute quantitative parameters of target goods according to customs declaration information, and display the risk attribute quantitative parameters on the central control platform; The risk attribute quantitative parameters include risk category and risk degree. The risk category includes abnormal goods name, abnormal goods quantity, abnormal goods specification and abnormal goods origin. The risk degree is the weight of the risk category corresponding to the risk category. Transmit the customs declaration information to the classification model and output the quantitative parameters of the risk attributes of the target goods; and displaying the risk attribute quantification parameters on the central control platform; The training process of the classification model specifically includes: obtaining a preliminary training data set, performing feature engineering on the preliminary training data set, obtaining risk attribute quantification parameters, then labeling each data sample, generating a risk label, obtaining an available training data set and an available test data set, and training the classification model through the available training data set; when the classification model reaches a preset training condition, stopping the training, taking the classification model obtained by the last training as the trained classification model, and outputting it; The loading and unloading robot includes a chassis, a vehicle body control box, a hybrid robotic arm, a grasping end, a visual perception module and a control module; The chassis is located at the bottom of the loading and unloading robot and is used to drive the loading and unloading robot to move; The vehicle body control box is fixedly arranged on the chassis and is used to load the control module and the visual perception module; The hybrid robot arm is arranged on the chassis and is used to drive the grasping end to move in space; The grabbing end is arranged at the end of the hybrid robot arm, and is used to absorb and transfer the target goods; The visual perception module is arranged on the hybrid robotic arm, and is used to obtain the grasping posture of the target cargo; The control module is arranged on the vehicle body control box, and is used to control the chassis, the hybrid mechanical arm and the grabbing end to absorb the target cargo and then transfer it to the conveying device; Wherein, the control module is electrically connected to the chassis, the visual perception module, the hybrid robot arm and the grasping end respectively; The gripping end includes a base, a vacuum suction cup and a conveyor belt; The bottom bracket is used to support the vacuum suction cup and the conveyor belt; The vacuum suction cup is slidably arranged on the bottom support through a slider, and is used to absorb the target goods onto the conveyor belt; The conveyor belt is covered on the bottom support and is used to transport the target goods to the conveying device; At the end of the gripping process, the vertical movement freedom of the vacuum suction cup adopts a parallel four-bar linkage combined with a cam guide mechanism that sets the lifting trajectory, and the horizontal movement freedom of the vacuum suction cup adopts a linear guide mechanism. The movement of the two is completed by the same motor driving the linear guide slider; The visual perception module includes a laser radar depth camera and a neural network. The laser radar obtains the depth information of the object being measured and obtains the grasping posture of the target cargo. The random sampling consensus algorithm is used to extract the grasping plane of the packaging box, and the local points belonging to the plane are used as the final valid point set. The algorithm for calculating the minimum area enclosing rectangle of the point set is used to obtain the rectangular boundary of the top surface of the cargo grasping. The control module controls the chassis, the hybrid robot arm and the grasping end to absorb the target cargo and transfer it to the conveying device through the grasping posture of the target cargo and the rectangular boundary of the top surface of the cargo grasping.

2. The integrated intelligent inspection system for boxed cargo according to claim 1 is characterized in that: The conveying device comprises a frame, a back-rolling conveyor belt, a lifting module and a roller conveyor belt; The frame is used to fix the back-rolling conveyor belt, the lifting module and the roller conveyor belt; The back roller conveyor belt and the lifting module are slidably arranged on the frame, the roller conveyor belt is detachably arranged on the frame, the roller conveyor belt and the lifting module are arranged on the same side of the frame, wherein the back roller conveyor belt is arranged on the opposite side of the roller conveyor belt and the lifting module; The back-rolling conveyor belt is used to place the target goods; The lifting module is used to adjust the height of the back-rolling conveyor belt to the height of the roller conveyor belt; The roller conveyor belt is used to convey the target goods to the inspection device.

3. The integrated intelligent inspection system for boxed cargo according to claim 1 is characterized in that: The inspection device includes a conveying module, a weighing module, a 3D camera, a control cabinet and an audible and visual alarm module; The conveying module is used to transfer the target goods to the weighing module; The weighing module is disposed below the conveying module, and is used to measure the weight of the target goods, obtain weighing information, and transmit the weighing information to the control cabinet; The 3D camera is fixedly mounted on the control cabinet, and is used to measure the size and count the quantity of the target goods, obtain size information and quantity information, and transmit the size information and quantity information to the control cabinet; The control cabinet is arranged beside the conveying module, and is used to obtain the measurement result according to the weighing information, the quantity information and the size information, compare the parameter information with the measurement result to obtain the inspection result, judge whether the target goods are abnormal according to the inspection result, obtain a first judgment result, and upload the first judgment result to the central control platform; The sound and light alarm module is arranged on the control cabinet and is used to determine whether to give an alarm according to the inspection result; Wherein, the control cabinet is electrically connected to the conveying module, the weighing module, the 3D camera and the sound and light alarm module respectively.

4. The integrated intelligent inspection system for boxed cargo according to claim 1 is characterized in that: The mark recognition device includes a mark recognition frame, an industrial camera, a motion mechanism, a rotating platform and a mark recognition conveying module; The mark recognition frame is used to fix the rotating platform and the motion mechanism; The motion mechanism is arranged on the mark recognition frame and is used to rotate the target goods; The industrial camera is used to take an initial photograph of the target goods to obtain an initial position photograph, control the motion mechanism to move to the mark position according to the initial position photograph, take a photograph of the mark of the target goods to obtain a mark position photograph, extract the mark information of the target goods according to the mark position photograph, compare the pre-acquired customs declaration information with the mark information, judge whether the target goods are abnormal, obtain a second judgment result, and upload the second judgment result and the mark information to the central control platform; The movement mechanism is arranged on the mark recognition frame and is used to move according to the initial position photo; The mark recognition and conveying module is arranged on the mark recognition frame, and is used for conveying the target goods to the separation device.

5. The integrated intelligent inspection system for boxed cargo according to claim 1 is characterized in that: The separation device includes a roller conveyor belt, a rotating mechanism and an information transmission device; The roller conveyor belt is connected to the rotating mechanism and is used to convey the target goods to the target location; The rotating mechanism is used to rotate according to the judgment result; The information transmission device is built into the rotating mechanism, and is used to receive the judgment result generated by the central control platform according to the first judgment result and the second judgment result, obtain the target position information, and push the target position information to the central control platform.

6. An inspection and processing method based on the integrated intelligent inspection system for boxed cargo according to any one of claims 1 to 5, characterized in that: The inspection and processing method comprises: The central control platform obtains the inspection instruction, generates an execution action instruction according to the inspection instruction, and the loading and unloading robot grabs the target goods according to the execution action instruction and places the target goods on the conveying device. At the same time, the central control platform generates risk attribute quantification parameters of the target goods according to the customs declaration information obtained in advance, and displays the risk attribute quantification parameters on the central control platform; The conveying device transmits the target goods to the inspection device, the inspection device performs size measurement, quantity counting and weight measurement on the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result, obtains a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device; The mark recognition device is connected to the end of the inspection device, and is used to extract the mark information of the target goods, compare the pre-acquired customs declaration information with the mark information, determine whether the target goods are abnormal, obtain a second judgment result, upload the second judgment result and the mark information to the central control platform, and transport the target goods to the separation device; The separation device receives a judgment result generated by the central control platform according to the first judgment result and the second judgment result, transports the target goods to a target location according to the judgment result, and pushes the target location information to the central control platform.

7. The inspection processing method according to claim 6, characterized in that: The inspection device measures the size, quantity and weight of the target goods to obtain the measurement result, compares the preset parameter information with the measurement result to obtain the inspection result, judges whether the target goods are abnormal according to the inspection result to obtain a first judgment result, uploads the first judgment result to the central control platform, and transports the target goods to the mark recognition device, specifically including: The conveying module transfers the target goods to the weighing module; The weighing module measures the weight of the target cargo to obtain weighing information, and transmits the weighing information to the control cabinet; The 3D camera measures the size and counts the quantity of the target goods to obtain size information and quantity information, and transmits the size information and quantity information to the control cabinet; The control cabinet obtains the measurement result according to the weighing information, the quantity information and the size information, compares the parameter information with the measurement result to obtain the inspection result, determines whether the target cargo is abnormal according to the inspection result, obtains a first judgment result, determines whether to alarm according to the first judgment result, uploads the first judgment result to the central control platform, and transports the target cargo to the mark recognition device.

8. The inspection processing method according to claim 6, characterized in that: The central control platform generates risk attribute quantification parameters of target goods according to the pre-acquired customs declaration information, and displays the risk attribute quantification parameters on the central control platform, specifically including: Inputting the customs declaration information into the trained classification model; Generate risk attribute quantified parameters of target goods through the classification model, and display the risk attribute quantified parameters on the central control platform; Wherein, the classification model is trained based on historical inspection data.

Images