ETC (Electronic Toll Collection) type digital intelligent arrival goods acceptance method, device and system

Abstract

The invention relates to the technical field of RFID (Radio Frequency Identification), in particular to an ETC (Electronic Toll Collection) type digital intelligent arrival goods acceptance method, device and system, and the method comprises the steps: transmitting a card reading signal through an RFID module, reading an RFID tag on a vehicle, and obtaining tag information; the acceptance module compares the goods acceptance information with the goods information to generate first acceptance information; an image recognition camera obtains a vehicle-mounted cargo image, image recognition and visual positioning are carried out, and a positioning packaging box set and a positioning packaging box label set are obtained; the RFID module transmits a positioning signal, positions the RFID tags of the packaging box and the goods based on the signal characteristics of the reflected signal, and obtains the position information of each tag; and the acceptance module generates the position arrangement of the packaging box and the association relationship between the packaging box and the goods based on the position information of each label, generates second acceptance information based on the position arrangement, and generates third acceptance information based on the association relationship. According to the method, the acceptance efficiency is greatly improved, and efficient acceptance, checking and sub-library storage are facilitated.

Description

Technical Field

[0001] The present application belongs to the field of RFID radio frequency identification technology, and specifically relates to an ETC-style digital arrival acceptance method, device and system. Background Art

[0002] In the supply chain fulfillment execution arrival and inspection business link, suppliers transport materials to project construction sites such as substations and material stations. The handover and acceptance of materials have the following difficulties. First, the construction sites are numerous and widespread, and the fulfillment personnel cannot go to multiple project sites at the same time, which is time-consuming and labor-intensive. Second, the number of arrived materials must be manually counted and the actual objects must be checked one by one, which is a large workload and inefficient. Third, the acceptance process requires manual recording, which is prone to errors, increasing the risks of material exchange, missing arrival quantity, and false arrival.

[0003] Specifically, when it comes to the execution phase of acceptance, inventory and warehousing, the packaging boxes need to be opened to identify, count and count the goods. For large projects, different items need to be stored in separate warehouses, which further increases the acceptance time and reduces the acceptance efficiency. Summary of the invention

[0004] The present application provides an ETC-style digital arrival acceptance method, device and system to solve or partially solve the problems raised in the above-mentioned background technology.

[0005] This application provides an ETC-style digital arrival acceptance method, including the following steps:

[0006] S1: The RFID module transmits a card reading signal to read the RFID tag on the vehicle and obtain tag information, which includes vehicle information, packaging box information and product information;

[0007] S2: The acceptance module queries the goods acceptance information according to the vehicle information, compares the goods acceptance information with the goods information, generates first acceptance information, and transmits the packaging box information and the goods information to the image recognition camera;

[0008] S3: The image recognition camera obtains the image of the cargo on the vehicle, performs image recognition and visual positioning on the image of the cargo on the vehicle according to the packaging box information, and obtains a set of positioned packaging boxes and a set of positioned packaging box labels;

[0009] S4: The RFID module transmits a positioning signal, locates the RFID tags of the packaging box and the goods based on the signal characteristics of the reflected signal, and obtains the location information of each tag;

[0010] S5: The acceptance module generates a position arrangement of the packaging boxes and an association relationship between the packaging boxes and the goods based on the position information of each tag, generates second acceptance information based on the corresponding position arrangement, and generates third acceptance information based on the corresponding association relationship.

[0011] Preferably, in the step S1, the vehicle information at least includes a vehicle identification code and a license plate number, the packing box information at least includes a packing box identification code and packing box features, the goods information at least includes a goods identification code, the vehicle information is obtained through a tag provided on the vehicle, the packing box information is obtained through a plurality of positioning tags provided on the outer end face of each packing box, and the goods information is obtained through a tag provided on the goods;

[0012] The shape of each packing box is a cuboid, and the pose of the corresponding packing box can be determined based on the positions of the positioning tags.

[0013] Preferably, in the step S3, the specific methods for the image recognition camera to obtain the set of positioned packing boxes, the set of positioned packing box tags, and the set of unpositioned packing boxes are as follows:

[0014] S301: Segment the vehicle picture to obtain the vehicle cargo stack image, perform image recognition on the cargo stack image based on the packing box features, recognize each packing box and perform visual positioning on it, and generate a set of positioned packing boxes {(packing box identification code, packing box pose)};

[0015] S302: Perform image recognition and visual positioning on the RFID tags on the recognized packing boxes to generate a set of positioned packing box tags.

[0016] Preferably, in the step S4, the specific method for the RFID module to obtain the position information of each tag is as follows:

[0017] S401: Use the RFID module to establish a position fingerprint model based on the received signal strength indication (RSSI) and phase in the bayonet area;

[0018] S402: Use the actual positions of the positioning tags in the set of positioned packing box tags to correct the position fingerprint model;

[0019] S403: Detect the RSSI values and phases of the reflected signals of the packing boxes and the tags to be positioned on the goods except for the elements in the set of positioned packing box tags, and match them with the corrected position fingerprint model to obtain the positions of each tag to be positioned.

[0020] Preferably, in the step S401, the reference tags in the position fingerprint model include real tags and virtual tags, and the insertion method of the virtual tags is as follows:

[0021] S411: Obtain the interpolation point coordinates c and the set of real tag coordinates {pi}, i = 1, 2,...., N, and select the Gaussian radial basis function as the kernel function

[0022]

[0023] Among them, " / / " represents the Euclidean norm, which is the attenuation coefficient and is usually set to 1 / (2σ 2 ), where σ is the signal standard deviation;

[0024] S412: Construct an interpolation equation to obtain the signal eigenvalue v(c) of the insertion point

[0025]

[0026] Among them, λ i is the weight coefficient, β is the bias term, and additional constraint conditions are added to ensure the uniqueness of the solution

[0027]

[0028] S413: Convert the interpolation equation and the constraint conditions into matrix form for solution

[0029]

[0030] Among them, Φ ∈ R NxN , which is the radial basis function matrix, and the element Φ ij = φ( / / p i -p j / / ), 1 ∈ R N×1 , 1 T ∈ R 1×N , λ ∈ R N×1 , which is the weight coefficient matrix, λ = [λ1, λ2,..., λ N T , V ∈ R N×1 , which is the characteristic signal matrix of the true label, V = [v1, v2,..., v N T ;

[0031] Solve the linear equations to obtain λ i and β;

[0032] S414: Associate the obtained signal eigenvalue v(c) with the coordinate c.

[0033] Preferably, in the step S402, the positioning packaging box label set is set as C1 = {pi}, where i is the label number. The specific method for correcting the position fingerprint model using the actual positions of the RFID tags in C1 is as follows:

[0034] S421: Use the position fingerprint model to fit the parameter curves F1(p) and F2(p), and set the parameter curves of the corrected position fingerprint model as G1(p) and G2(p). The correction function is set as:

[0035] ​​M1(p) = G1(p) / F1(p), M2(p) = G2(p) / F2(p) (1)

[0036] Where p is the position coordinate, the values of F1(p) and G1(p) are the RSSI values at each coordinate, and the values of F2(p) and G2(p) are the phase angles at the corresponding coordinates;

[0037] S422: Calculate the theoretical signal characteristics of each position according to the actual positions of the tags in set C1 and the position fingerprint model, and generate set C2 = {(M1(pi), M2(pi))};

[0038] S423: Use the RFID module to actually detect the actual signal characteristics at the positions of the tags in set C1, generate set C3 = {(G1(pi), G2(pi))}, and according to formula (1), further obtain the values of the correction functions M1(pi) and M2(pi), and perform fitting on M1(p) and M2(p) based on set C1 and the values of M1(pi) and M2(pi);

[0039] S424: Use M1(p) and M2(p) to correct the signal characteristics of each reference tag in the position fingerprint model.

[0040] Preferably, in step S422, the method of calculating the theoretical signal characteristics of each position according to the actual positions of the tags in set C1 and the position fingerprint model can be implemented according to the method of inserting virtual tags. Obtain the coordinate d of a certain tag, and obtain the signal characteristics of the reference tag adjacent to position d in the position fingerprint model to generate the signal characteristics at position d;

[0041] Set up two acceptance devices with the same RFID module in the same area to construct a vehicle checkpoint. In step S422, each point in set C1 constitutes two position coordinates and two sets of signal characteristics with respect to the two RFID modules. The two acceptance devices communicate with each other, and the elements of set C2 can be doubled.

[0042] Preferably, in step S403, the specific method of detecting the signal characteristics of the signal of the tag to be located and matching with the corrected position fingerprint model to obtain the positions of each tag to be located is as follows:

[0043] S431: Normalize the signal characteristics of the tag to be located and the signal characteristics (RSSI intensity and phase) of each reference tag in the position fingerprint model to eliminate the dimensional difference;

[0044] S432: Use the Gaussian kernel function to calculate the similarity weights between the tag to be located and the reference tags

[0045]

[0046] Among them, Var is to calculate the variance, and V norm ∈R N×2 is the signal feature matrix of each reference tag after normalization, ΔRSSI is the RSSI difference between the tag to be located and the reference tag i, and Δθ is the phase difference between the tag to be located and the reference tag i;

[0047] S433: Select the K reference tags with the highest similarity weights and perform weighted positioning on the tags to be located:

[0048]

[0049] Among them, (xi, yi) is the coordinate of the tag numbered i.

[0050] This application also provides an ETC-based digital intelligent arrival acceptance device, including: an image recognition camera and an RFID reader, and also including a controller and an Internet of Things module, which can implement the ETC-based digital intelligent arrival acceptance method as described above.

[0051] This application also provides an ETC-based digital intelligent arrival acceptance system, including: a cloud platform and several ETC-based digital intelligent arrival acceptance devices as described above.

[0052] Compared with the prior art, the beneficial effects of this application are:

[0053] (1) Through the mutual cooperation of RFID radio frequency identification technology and image recognition technology, this application identifies and locates the goods and packaging boxes, and then automatically generates the first acceptance information related to the types and quantities of goods, the second acceptance information related to the layout of packaging boxes, and the third acceptance information related to the matching of packaging boxes and goods, greatly improving the acceptance efficiency and facilitating the efficient progress of acceptance, inventory, and storage in different warehouses.

[0054] (2) This application locates the packaging boxes on the contour of the vehicle cargo stack through image recognition and positioning, locates the packaging boxes inside the cargo stack through RFID radio frequency identification technology, and calibrates the RFID positioning through the high-precision positioning of image recognition, effectively realizing the accurate positioning of the packaging boxes and goods at all positions of the cargo stack.

[0055] (3) This application uses the RSSI value and phase two-dimensional parameter features to identify the spatial distribution of reflected signals, effectively avoiding the inaccuracy of single RSSI value positioning and facilitating the guarantee of positioning accuracy in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The following further illustrates this application with reference to the drawings and embodiments.

[0057] Figure 1 is a schematic diagram of the device structure of this application,

[0058] Figure 2 It is a schematic diagram of the system implementation of this application.

[0059] Figure 3 It is a schematic diagram of the method flow of this application.

[0060] In the figure:

[0061] 1. Installation base column, 2. Image recognition camera, 3. Buzzer, 4. Indicator light, 5. RFID card reader, 6. Monitor, 7. Infrared induction switch, 8. Base;

[0062] 100. Acceptance device body, 200. Vehicle to be accepted. Specific implementation manners

[0063] As used in the specification and claims, certain terms are used to refer to specific components. Those skilled in the art should understand that hardware manufacturers may use different terms to refer to the same component. The specification and claims do not use the difference in names as a way to distinguish components, but use the difference in functions of components as the criterion for distinction. As mentioned throughout the specification and claims, "comprising" is an open-ended term and should be interpreted as "including but not limited to". "Roughly" means within an acceptable error range. Those skilled in the art can solve the technical problem within a certain error range and basically achieve the technical effect.

[0064] In the description of this application, it should be understood that the orientation or positional relationship indicated by the terms "upper", "lower", "front", "rear", "left", "right", "horizontal", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing this application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0065] In this application, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific situations.

[0066] Embodiment 1

[0067] As Figures 1 to 3 shown, this application provides an ETC-style digital intelligent arrival acceptance method, which specifically includes the following steps:

[0068] S1: The RFID module emits a card-reading signal to read the RFID tag on the vehicle and obtain the tag information, where the tag information includes vehicle information, packing box information, and goods information;

[0069] S2: The acceptance module queries the goods acceptance information according to the vehicle information, compares the goods acceptance information with the goods information to generate the first acceptance information, and transmits the packing box information and the goods information to the image recognition camera;

[0070] S3: The image recognition camera obtains the on-vehicle cargo image, performs image recognition and visual positioning on the on-vehicle cargo image according to the packing box information, and obtains the positioned packing box set and the positioned packing box tag set;

[0071] S4: The RFID module emits a positioning signal, positions the RFID tags of the packing boxes and the goods based on the signal characteristics of the reflected signal, and obtains the position information of each tag;

[0072] S5: The acceptance module generates the position arrangement of the packing boxes and the association relationship between the packing boxes and the goods based on the position information of each tag, generates the second acceptance information based on the corresponding position arrangement, and generates the third acceptance information based on the corresponding association relationship.

[0073] Specifically, in the step S1, the vehicle information at least includes a vehicle identification code and a license plate number, the packing box information at least includes a packing box identification code and packing box characteristics, the goods information at least includes a goods identification code, the vehicle information is obtained through a tag provided on the vehicle, the packing box information is obtained through a number of positioning tags provided on the outer end face of each packing box, and the goods information is obtained through a tag provided on the goods.

[0074] Furthermore, the shape of each packing box is a cuboid, and a setting position for an RFID-based positioning tag is reserved on the outer end face of each packing box. The pose of the corresponding packing box can be determined based on the position of the positioning tag. The packing box characteristics include the packing box size, the positioning tag position, and the external pattern and text characteristics of the packing box.

[0075] Specifically, in the step S2, the acceptance module communicates with the cloud platform to obtain the goods acceptance information corresponding to the vehicle information. The goods acceptance information at least includes the goods name (type) and the corresponding goods name identification code and the quantity of each type. The goods name identification code and the corresponding goods name can be determined through the goods identification code.

[0076] Specifically, in the step S3, the specific methods for the image recognition camera to obtain the positioned packing box set, the positioned packing box tag set, and the un-positioned packing box set are as follows:

[0077] S301: Segment the vehicle images to obtain the vehicle cargo stack images. Based on the characteristics of the packaging boxes, perform image recognition on the cargo stack images, identify each packaging box and perform visual positioning on it, and generate a set of positioned packaging boxes {(packaging box identification code, packaging box pose)}.

[0078] S302: Perform image recognition and visual positioning on the RFID tags on the recognized packaging boxes to generate a set of positioned packaging box tags.

[0079] Steps S301 to S303 are implemented using an image recognition module built into the image recognition camera. This application does not involve improvements to the image recognition algorithm and the visual positioning algorithm, and the algorithm principle will not be elaborated here. An OCR module is introduced into the image recognition module to extract the text on the outer packaging of the goods to enhance the recognition and discrimination ability for outer packages with similar shapes.

[0080] In step S4, the specific method for the RFID module to obtain the position information of each tag is as follows:

[0081] S401: Use the RFID module to establish a position fingerprint model based on the received signal strength indication (RSSI) and phase in the bayonet area.

[0082] S402: Use the actual positions of the positioned tags in the set of positioned packaging box tags to correct the position fingerprint model.

[0083] S403: Detect the RSSI value and phase of the reflected signals of the packaging boxes and the tags to be positioned on the goods other than the elements in the set of positioned packaging box tags, and match them with the corrected position fingerprint model to obtain the positions of each tag to be positioned.

[0084] In traditional RFID positioning methods, the received signal strength indication (RSSI) is mainly used as the positioning parameter, which has insufficient stability and poor positioning accuracy in complex environments. In this application, phase parameters are introduced together with RSSI values to perform two-dimensional characterization of the reflected signal characteristics, strengthening the accuracy of signal characteristic characterization in complex environments and facilitating the improvement of positioning accuracy.

[0085] In step S401, the accuracy of the position fingerprint model depends on the density of the reference tags. However, obviously, too many reference tags will increase the cost and test complexity. The common practice in the industry is to introduce virtual reference tags, that is, insert virtual reference tags between real tags to increase the density of the reference tags. The insertion of virtual reference tags can be linear insertion or non-linear insertion.

[0086] Specifically, the method of linear insertion is as follows:

[0087] For two adjacent real tags A(Pa , V a ), and B(P b , V b ), the value V of the signal feature of the virtual tag C inserted on the line connecting them AB c is:

[0088]

[0089] where P a and P b are the coordinates of points A and B, V a and V b are the values of the signal features at positions A and B, that is, the phase angle or the RSSI value. The phase and RSSI values at the virtual tag C are calculated respectively through the above formula.

[0090] Obviously, the generation of virtual tags in linear interpolation depends on the adjacent real tags at both ends. The generation method is simple, but the accuracy is insufficient in complex environments. More real tags can be introduced as the generation reference for virtual tags.

[0091] As a preferred technical solution, the present application introduces Radial Basis Function (RBF) interpolation as the insertion method for virtual tags. The specific steps are as follows:

[0092] S411: Obtain the interpolation point coordinates c and the set of real tag coordinates {p i}, i = 1, 2,...., N, and select the Gaussian radial basis function as the kernel function

[0093]

[0094] where, " / / " represents the Euclidean norm, is the attenuation coefficient, usually set = 1 / (2σ 2 ), and σ is the signal standard deviation;

[0095] S412: Construct an interpolation equation to obtain the signal feature value v(c) of the insertion point

[0096]

[0097] where λ i is the weight coefficient, β is the bias term, and additional constraint conditions are added to ensure the uniqueness of the solution

[0098]

[0099] S413: Convert the interpolation equation and the constraint conditions into matrix form for solution

[0100]

[0101] where, Φ ∈ R NxN , is a radial basis function matrix, and the element Φ ij = φ( / / p i - p j / / ), 1 ∈ R N×1 , 1 T ∈ R 1×N , λ ∈ R N×1 , is a weight coefficient matrix, λ = [λ1, λ2,..., λ N T , V ∈ R N×1 , is a characteristic signal matrix of the true label, V = [v1, v2,..., v N T ;

[0102] Solve the linear equations to obtain λ i and β;

[0103] S414: Associate the obtained signal eigenvalue v(c) with the coordinate c.

[0104] In the step S413, the linear equations can be solved by Cholesky decomposition, conjugate gradient method or other linear algebra libraries.

[0105] Based on steps S411 to S414, establish the correlation relationships between the RSSI values, phase angles and the inserted coordinates respectively, and then generate a location fingerprint model.

[0106] In step S411, when the number of true labels is sufficient, select the N true labels closest to the interpolation point coordinate c.

[0107] In the step S402, set the positioning packaging box label set as C1 = {pi}, where i is the label number. The specific method for correcting the location fingerprint model using the actual positions of the RFID tags in C1 is as follows:

[0108] S421: Use the location fingerprint model to fit the parameter curves F1(p) and F2(p), set the parameter curves of the corrected location fingerprint model as G1(p) and G2(p), and the correction function is set as:

[0109] M1(p) = G1(p) / F1(p), M2(p) = G2(p) / F2(p) (1)

[0110] where, p is the location coordinate, the values of F1(p) and G1(p) are the RSSI values at each coordinate, and the values of F2(p) and G2(p) are the phase angles at the corresponding coordinates;

[0111] ​​S422: Calculate the theoretical signal features of each position according to the actual positions of the tags in set C1 and the position fingerprint model, and generate set C2 = {(M1(pi), M2(pi))};

[0112] S423: Use the RFID module to actually detect the actual signal features at the positions of the tags in set C1, generate set C3 = {(G1(pi), G2(pi))}, and according to formula (1), further obtain the values of the correction functions M1(pi) and M2(pi), and perform fitting on M1(p) and M2(p) based on set C1 and the values of M1(pi) and M2(pi);

[0113] S424: Use M1(p) and M2(p) to correct the signal features of each reference tag in the position fingerprint model.

[0114] In step S422, the method for calculating the theoretical signal features of each position according to the actual positions of the tags in set C1 and the position fingerprint model can be implemented according to the virtual tag insertion method described above. Obtain the coordinates d of a certain tag, obtain the signal features of the reference tags (which can be real tags or virtual tags) adjacent to position d in the position fingerprint model, and generate the signal features at position d. The radial basis function interpolation method is preferred.

[0115] Furthermore, as Figure 2 shown, in the actual application scenario, two acceptance devices carrying the same RFID module will be set in the same area to construct a vehicle checkpoint. Therefore, in step S422, each point in set C1 constitutes two position coordinates and two sets of signal features with respect to the two RFID modules. By communicating with each other, the elements of set C2 can be doubled, which is convenient for the accurate fitting of M1(p) and M2(p).

[0116] In step S403, the signal features (RSSI intensity and phase) of the signal of the tag to be located are detected and matched with the corrected position fingerprint model. The specific method for obtaining the positions of each tag to be located is as follows:

[0117] S431: Normalize the signal features of the tag to be located and the signal features (RSSI intensity and phase) of each reference tag in the position fingerprint model to eliminate the dimensional difference;

[0118] S432: Use the Gaussian kernel function to calculate the similarity weights between the tag to be located and the reference tags

[0119]

[0120] where, Var is to calculate the variance, V norm ∈R N×2is the signal feature matrix of each reference tag after normalization, ΔRSSI is the RSSI difference between the tag to be located and the reference tag i, and Δθ is the phase difference between the tag to be located and the reference tag i;

[0121] S433: Select the K reference tags with the highest similarity weights and perform weighted positioning on the tags to be located:

[0122]

[0123] where (x i , y i ) are the coordinates of the tag numbered i.

[0124] In the step S431, the typical normalization method is Z-Score normalization.

[0125] In the step S5, the specific methods for the acceptance module to generate the second acceptance information and the third acceptance information are as follows:

[0126] Generate the poses (positions and postures) of each packing box except the elements of the located packing box set according to the positions of the positioning tags of each packing box, combine the located packing box set to generate the overall arrangement of the packing boxes, match the goods to each packing box based on the positions of each goods tag, and then generate the association relationship between the packing boxes and the goods. Furthermore, generate the second acceptance information based on the corresponding position arrangement and the third acceptance information based on the corresponding association relationship.

[0127] Typically, the positioning tags of the packing boxes are set at each corner of a cuboid. The number of positioning tags (such as setting placement rules) can also be increased or decreased as long as the effect of determining the pose of the corresponding packing box can be achieved.

[0128] The first acceptance information is used to proofread the types and quantities of the goods. The second acceptance information and the third acceptance information are used to determine the association relationship between the packing boxes and the goods and the overall arrangement of the packing boxes, which is convenient for subsequent unpacking and separate storage of the goods.

[0129] Embodiment 2

[0130] The present application provides an ETC-based digital intelligent arrival acceptance device, including a base 8. An installation base column 1 is vertically provided on the base. On the same-side end faces of the installation base column 1, an image recognition camera 2 and an RFID card reader 5 are respectively provided. A controller and an Internet of Things module are provided inside the installation base column 1. The controller is electrically connected to the image recognition camera 2, the RFID card reader 5, and the Internet of Things module. The acceptance module described in Embodiment 1 is deployed inside the controller. The RFID card reader corresponds to the RFID module described in Embodiment 1. The acceptance module, the image recognition camera 2, and the RFID card reader 5 cooperate with each other to implement the ETC-based digital intelligent arrival acceptance method in Embodiment 1.

[0131] Preferably, a buzzer 3, an indicator light 4, a display 6, and an infrared induction switch 7 electrically connected to the controller are further provided on the installation base column.

[0132] The infrared induction switch 7 is used to detect the approach of a vehicle. When the vehicle approaches the preset distance of this device, the infrared induction switch 7 sends an alarm signal to the controller.

[0133] The buzzer 3 is used for audible and visual alarms when a vehicle approaches and when the acceptance is unqualified.

[0134] The indicator light 4 is used for lighting indication of whether the acceptance result is normal or not.

[0135] The display 6 facilitates viewing of the acceptance result.

[0136] The controller is an industrial computer device, and the Internet of Things module is a 5G wireless network module.

[0137] Embodiment 3

[0138] The present application provides an ETC-based digital intelligent arrival acceptance system, including a cloud platform and a number of acceptance device bodies 100 as described in Embodiment 2 that are wirelessly communicated with the cloud platform.

[0139] In each checkpoint area, two acceptance device bodies 100 are provided to accept the goods on the vehicle 200 to be accepted.

[0140] The above has described the embodiments of the present application in detail with reference to the accompanying drawings. However, the present application is not limited to the above embodiments. Within the knowledge scope of those of ordinary skill in the art to which the present application pertains, various changes can be made without departing from the purpose of the present application.

Claims

1. The ETC-based digital intelligent arrival inspection method is characterized in that It includes the following steps: S1: The RFID module emits a card-reading signal to read the RFID tag on the vehicle and obtain tag information, where the tag information includes vehicle information, packing box information, and goods information; S2: The acceptance module queries the goods acceptance information according to the vehicle information, compares the goods acceptance information with the goods information to generate the first acceptance information, and transmits the packing box information and the goods information to the image recognition camera; S3: The image recognition camera acquires the on-vehicle cargo image, performs image recognition and visual positioning on the on-vehicle cargo image according to the packing box information, and obtains a set of positioned packing boxes and a set of positioned packing box tags; S4: The RFID module emits a positioning signal, locates the RFID tags of the packing boxes and goods based on the signal characteristics of the reflected signal, and obtains the position information of each tag; S5: The acceptance module generates the position arrangement of the packing boxes and the association relationship between the packing boxes and the goods based on the position information of each tag, generates the second acceptance information based on the corresponding position arrangement, and generates the third acceptance information based on the corresponding association relationship.

2. The ETC-based digital intelligent arrival acceptance method according to claim 1, wherein: In the step S1, the vehicle information includes at least a vehicle identification code and a license plate number, the packing box information includes at least a packing box identification code and packing box characteristics, the goods information includes at least a goods identification code, the vehicle information is obtained through a tag provided on the vehicle, the packing box information is obtained through a number of positioning tags provided on the outer end face of each packing box, and the goods information is obtained through a tag provided on the goods; The shape of each packing box is a cuboid, and the pose of the corresponding packing box can be determined based on the position of the positioning tag.

3. The ETC-based digital intelligent arrival acceptance method according to claim 1, wherein: In the step S3, the specific method for the image recognition camera to obtain a set of positioned packing boxes, a set of positioned packing box tags, and a set of unpositioned packing boxes is as follows: S301: Segment the vehicle picture to obtain the vehicle cargo stack image, perform image recognition on the cargo stack image based on the packing box characteristics, identify each packing box and perform visual positioning on it to generate a set of positioned packing boxes {(packing box identification code, packing box pose)}; S302: Perform image recognition and visual positioning on the RFID tags on the identified packing boxes to generate a set of positioned packing box tags.

4. The ETC-based digital intelligent arrival acceptance method according to claim 1, wherein: In the step S4, the specific method for the RFID module to obtain the position information of each tag is as follows: S401: Use the RFID module to establish a position fingerprint model based on the received signal strength indication (RSSI) and phase of the reflected signal in the checkpoint area; S402: Use the actual positions of the positioning tags in the set of positioned packing box tags to correct the position fingerprint model; S403: Detect the RSSI value and phase of the reflected signal of the packing boxes and goods to be positioned except for the elements in the set of positioned packing box tags, and match them with the corrected position fingerprint model to obtain the positions of each to-be-positioned tag.

5. The ETC-based digital intelligent arrival acceptance method according to claim 4, wherein: In the step S401, the reference tags in the location fingerprint model include real tags and virtual tags. The method for inserting virtual tags is as follows: S411: Obtain the interpolation point coordinates c and the set of real tag coordinates {pi}, where i = 1, 2,...., N, and select the Gaussian radial basis function as the kernel function r = / / c-p i / / Among them, " / / " represents the Euclidean norm, ∈ is the attenuation coefficient, and usually ∈ = 1 / (2σ 2 ), where σ is the signal standard deviation; S412: Construct an interpolation equation to obtain the signal eigenvalue v(c) of the insertion point where λi is the weight coefficient and β is the bias term, and additional constraint conditions are added to ensure the uniqueness of the solution S413: Convert the interpolation equation and the constraint conditions into matrix form for solution where, Φ ∈ R NxN , is a radial basis function matrix, and the element φ ij = φ( / / p i -p j / / ), 1 ∈ R N×1 , 1 T ∈ R 1×N , λ ∈ R N ×1 , is a weight coefficient matrix, λ = [λ1, λ2,..., λ N T , V ∈ R N×1 , is a characteristic signal matrix of the true label, V = [v1, v2,..., v N T ;​​ Solve the linear equations to obtain λ i and β; S414: Associate the obtained signal eigenvalue v(c) with the coordinates c.

6. The ETC-based digital arrival inspection method according to claim 4, characterized in that: In the step S402, the set of positioned packaging box tags is set as C1 = {pi}, where i is the tag number. The specific method for correcting the location fingerprint model using the actual positions of the RFID tags in C1 is as follows: S421: Use the location fingerprint model to fit the parameter curves F1(p) and F2(p), and set the parameter curves of the corrected location fingerprint model as G1(p) and G2(p). The correction functions are set as: M1(p) = G1(p) / F1(p), M2(p) = G2(p) / F2(p) (1) where p is the position coordinate, the values of F1(p) and G1(p) are the RSSI values at each coordinate, and the values of F2(p) and G2(p) are the phase angles at the corresponding coordinates; S422: Calculate the theoretical signal characteristics of each position according to the actual positions of the tags in the set C1 and the location fingerprint model, and generate the set C2 = {(M1(pi), M2(pi))}; S423: Use the RFID module to actually detect the actual signal characteristics at the positions of the tags in the set C1, and generate the set C3 = {(G1(pi), G2(pi))}. According to equation (1), the values of the correction functions M1(pi) and M2(pi) are obtained, and M1(p) and M2(p) are generated by fitting based on the set C1 and the values of M1(pi) and M2(pi); S424: Use M1(p) and M2(p) to correct the signal characteristics of each reference tag in the location fingerprint model.

7. The ETC-based digital arrival inspection method according to claim 6, characterized in that: In the step S422, the method for calculating the theoretical signal characteristics of each position according to the actual positions of the tags in the set C1 and the location fingerprint model can be implemented according to the method for inserting virtual tags. Obtain the coordinates d of a certain tag, and obtain the signal characteristics of the reference tags at the adjacent positions d in the location fingerprint model to generate the signal characteristics at the position d; Two inspection devices equipped with the same RFID module are set in the same area to construct a vehicle checkpoint. In the step S422, each point in the set C1 constitutes two position coordinates and two sets of signal characteristics with respect to the two RFID modules. The two inspection devices communicate with each other, and the elements of the set C2 can be doubled.

8. The ETC-based digital arrival inspection method according to claim 4, characterized in that: In the step S403, the specific method for obtaining the positions of the to-be-located tags by matching the signal features of the detected to-be-located tags with the corrected location fingerprint model is as follows: S431: Normalize the signal features of the to-be-located tags and the signal features of each reference tag in the location fingerprint model to eliminate the dimensional difference; S432: Use the Gaussian kernel function to calculate the similarity weights between the to-be-located tags and the reference tags Among them, Var is to calculate the variance, V norm ∈R N×2 is the signal feature matrix of each reference tag after normalization, ΔRSSI is the RSSI difference between the tag to be located and the reference tag i, and Δθ is the phase difference between the tag to be located and the reference tag i; S433: Select the K reference tags with the highest similarity weights and perform weighted positioning on the to-be-located tags: where (xi, yi) are the coordinates of the tag numbered i.

9. The ETC-based digital intelligent arrival inspection device is characterized in that including: An image recognition camera (2) and an RFID reader (5), further including a controller and an Internet of Things module, capable of implementing the ETC-style digital intelligent arrival acceptance method described in any one of claims 1-8.

10. The ETC-based digital intelligent arrival inspection system is characterized in that, including: A cloud platform and a number of ETC-style digital intelligent arrival acceptance devices as described in claim 9.

Images