CN116720809A

CN116720809A - Intelligent recognition receiving method and intelligent recognition receiving device thereof

Info

Publication number: CN116720809A
Application number: CN202310576242.4A
Authority: CN
Inventors: 刘霞; 刘迪; 陈一梅; 杨潇潇; 杨文峰; 杨雅琴; 梁志松
Original assignee: GAC Honda Automobile Co Ltd
Current assignee: GAC Honda Automobile Co Ltd
Priority date: 2023-05-19
Filing date: 2023-05-19
Publication date: 2023-09-08

Abstract

The invention relates to the technical field of warehouse logistics management, in particular to an intelligent recognition receiving method and an intelligent recognition receiving device thereof, wherein the method comprises the following steps: shooting a container by a shooting piece to obtain a pallet truck container image, and then carrying out target detection on the pallet truck container image to obtain coordinate positions corresponding to all the labels and obtain the number of the labels; then, character recognition screening is carried out on the labels to obtain first box number information; the code reading shooting piece reads the two-dimensional code on the label, and then second box number information is obtained through processing; comparing the number of the two-dimensional codes with the number of the labels; if the number is equal, carrying out corresponding matching on the first box number information and the second box number information, carrying out random matching if the first box number information and the second box number information cannot be corresponding, then comparing the first box number information with the receiving plan, and if the first box number information and the second box number information are in a complete matching state, completing goods receiving; the number of the box numbers is unequal, the box number information which cannot be matched correspondingly is not matched, and the goods are not received. The invention can realize timely and efficient goods receiving and improve the goods receiving efficiency.

Description

Intelligent recognition receiving method and intelligent recognition receiving device thereof

Technical Field

The invention relates to the technical field of warehouse logistics management, in particular to an intelligent recognition receiving method and an intelligent recognition receiving device thereof.

Background

At present, the receiving treatment of the parts commodity flow collection provided by suppliers is mainly manual operation, operators need to scan the identification codes on all parts containers actually arrived at the goods one by one, the scanning results are uploaded to a background system, and the total arrived parts quantity is calculated and then compared with the planned arrived parts quantity every day. The existing goods receiving treatment can only check the number of the parts to be received after the receiving work of the daily part container is completed. The receiving mode has the advantages of large code scanning amount, repeated and complicated operation process, high labor intensity, long time consumption, easy fatigue generation, incapability of effectively ensuring precision, possible missing scanning and error scanning, and low overall operation efficiency; when the load reaches the peak time, the situation that the parts stay and accumulate in the unloading area easily occurs, and the unloading operation is influenced; if a difference is found in the data in the subsequent verification operation, the post-trace work is large and the timeliness is poor. In the prior art, some intelligent warehouse management methods and management systems exist, storage management is carried out through FRID (fast Fourier transform) or computer vision detection and other modes, but the existing warehouse management mostly needs one-to-one complete matching of goods and manifest and can carry out warehouse management, and when a system identification error is encountered, goods accumulation easily occurs, so that goods cannot be received in time, and the receiving efficiency is inconvenient to improve.

Disclosure of Invention

The invention aims to overcome the defects of the prior art, and provides an intelligent recognition receiving method and an intelligent recognition receiving device thereof, which can realize timely and efficient receiving, avoid goods accumulation and improve receiving efficiency.

In order to solve the technical problems, the invention adopts the following technical scheme:

the intelligent identifying and receiving method comprises the following steps:

s1, after a cargo pallet to be signed passes through a triggering mechanism, triggering a shooting piece to shoot cargo boxes placed on the cargo pallet and obtain a pallet box image of the pallet, wherein the pallet box image of the pallet contains all cargo boxes on the current cargo pallet, labels are adhered to all cargo boxes, and the labels contain two-dimensional code information and character string information;

s2, transmitting the trolley container image to a processing control mechanism, and carrying out target detection on a plurality of labels in the trolley container image to obtain corresponding coordinate positions of the labels in the trolley container image and the number of the labels contained in the trolley container image; then, the processing control mechanism carries out OCR (optical character recognition) on the character string information of the tag to obtain a plurality of corresponding first box number information;

S3, after the goods carrying vehicle to be signed passes through the triggering mechanism, the code reading shooting piece is triggered to read the two-dimensional code on the label, and corresponding character information is obtained; the character information is transmitted to the processing control mechanism for screening, so that a plurality of corresponding second box number information is obtained, and the number of the two-dimension codes on the current cargo pallet is obtained;

s4, comparing the number of the two-dimensional codes with the number of the labels; if the number is equal, executing the step S5, otherwise executing the step S6;

s5, carrying out matching processing on the first box number information and the second box number information: carrying out one-to-one correspondence matching on the box number information which can be corresponding, carrying out random matching on the rest of box number information which cannot be corresponding, then summarizing and packaging container matching data of the current cargo pallet, and then executing step S7;

s6, carrying out matching processing on the first box number information and the second box number information: carrying out one-to-one correspondence matching on box number information which can be corresponding, carrying out no matching on the rest of box number information which cannot be corresponding, marking the box number information which cannot be matched in the cargo box image of the cargo box according to the corresponding coordinate position of the box number information, then summarizing and packaging cargo box matching data of the current cargo box, and then executing step S7;

S7, comparing the container matching data which is completed and packed with the current day receiving plan, if the current container matching data is in the current day receiving plan and the container matching data is in a matching completed state, completing container receiving of the current cargo pallet, otherwise, not receiving and marking the current cargo pallet.

Further, the triggering mechanism comprises a first triggering component, a second triggering component and a third triggering component which are sequentially arranged along the conveying direction of the cargo trolley; the step S1 specifically comprises the following steps: the processing control mechanism receives the first in-place signal, the second in-place signal and the third in-place signal of the cargo truck detected by the first triggering component, the second triggering component and the third triggering component, judges whether the receiving sequence of the first in-place signal and the second in-place signal is reverse, if not, triggers the shooting piece to take a picture, otherwise does not take a picture; the step S3 specifically includes the following steps: the processing control mechanism receives the first in-place signal, the second in-place signal and the third in-place signal of the cargo truck detected by the first triggering component, the second triggering component and the third triggering component, judges whether the receiving order of the first in-place signal and the second in-place signal is reverse, if not, triggers the code reading shooting piece to read the code, otherwise does not read the code; and the steps S1 and S3 are not sequential.

Further, the processing control mechanism comprises a target detection model and a character recognition model; the step S2 specifically includes the following steps: transmitting the pallet truck container image to the target detection model for target detection, obtaining coordinate positions corresponding to all labels in the pallet truck container image, and obtaining the number of labels contained in the pallet truck container image; and then, using the character recognition model to recognize character string information in the tag, and screening out characters conforming to the form of numbers to obtain a plurality of corresponding second box number information.

Further, the target detection model takes YOLOV5 as a basic technical frame, corresponding characteristics of the labels are extracted in a neural network by utilizing a multi-layer network, classification is carried out by scoring, and then the coordinate positions of the labels are positioned; the character recognition model is a CRNN character recognition model, text regions in the labels are extracted through det text detection, and character string information in the text regions is recognized and screened through rec text recognition.

Further, the display device further comprises a display mechanism, wherein:

the step S5 specifically includes the following steps: matching the first box number information with the second box number information: carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a first color wire frame in the display mechanism according to the corresponding coordinate position of the label; the rest can not be matched randomly, a second color wire frame is displayed in the display mechanism according to the corresponding coordinate position of the label, and then the goods receiving is completed;

The step S6 specifically includes the following steps:

s61, judging whether the first box number information has repeated information or not, judging whether the second box number information has repeated box number information or not, if so, executing the step S62, otherwise, executing the step S65;

s62, judging whether the repeated box number information part has a part which can be correspondingly matched, if so, executing a step S63, otherwise, executing a step S64;

s63, selecting one piece of repeated box number information to be correspondingly matched with the other piece of box number information which can be correspondingly matched for the part with the repeated information, and displaying a fourth color line frame in the display mechanism according to the corresponding coordinate position of the label; the rest repeated box number information is not matched, and a fourth color wire frame is also displayed in the display mechanism according to the corresponding coordinate position of the box number information which is not matched;

for the part with non-repeated information, matching the box number information which can be corresponding one by one, and displaying a first color wire frame in a display mechanism according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in a display mechanism according to the corresponding coordinate position of the box number information which cannot be matched, and then step S7 is executed;

S64, displaying a third color wire frame in the display mechanism according to the corresponding coordinate position of the part where the information is repeated; for the part with non-repeated information, matching the box number information which can be corresponding one by one, and displaying a first color wire frame in a display mechanism according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in a display mechanism according to the corresponding coordinate position of the box number information which cannot be matched, and then step S7 is executed;

s65, carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a first color line frame in a display mechanism according to the corresponding coordinate position of the label; and if the other box numbers cannot be matched correspondingly, the box number information which is not matched is displayed with a third color line frame in a display mechanism according to the corresponding coordinate position of the box number information, and then step S7 is executed.

The invention also provides an intelligent recognition receiving device, which comprises a code reading mechanism, a triggering mechanism and a processing control mechanism, wherein the code reading mechanism comprises a portal and a plurality of code reading shooting pieces and shooting pieces which are connected to the portal, and shooting ends of the code reading shooting pieces and the shooting pieces are both used for facing to a cargo carrying vehicle; the triggering mechanism comprises a plurality of sensing assemblies arranged along the conveying direction of the cargo pallet; the code reading shooting piece, the shooting piece and the sensing component are all in communication connection with the processing control mechanism.

Preferably, a first connecting rod is arranged in the portal frame, the code reading shooting piece and the shooting piece are connected to the first connecting rod in a sliding manner, and locking structures are arranged between the code reading shooting piece and the first connecting rod and between the shooting piece and the first connecting rod; the first connecting rod is vertically arranged, at least two code reading shooting pieces are arranged along the first connecting rod, the code reading shooting pieces are code reading cameras, and the shooting pieces are industrial cameras.

Preferably, the intelligent recognition receiving device further comprises a guide mechanism, wherein the guide mechanism comprises a plurality of guide rails arranged along the conveying direction of the cargo pallet, and the leading-in ends of the guide rails are provided with leading-in structures; the plurality of sensing assemblies are sequentially arranged on the guide rail along the extending direction of the guide rail; the number of the code reading mechanisms is two, and the guide rail is positioned between the two code reading mechanisms.

Preferably, the intelligent recognition receiving device further comprises a light supplementing mechanism which is in communication connection with the processing control mechanism, wherein a light source emission end of the light supplementing mechanism is used for facing the cargo carrying vehicle; the light supplementing mechanism comprises a second connecting rod and a plurality of light supplementing lamps movably connected to the second connecting rod, and the light supplementing lamps are in communication connection with the processing control mechanism through a light source controller.

Preferably, the processing control mechanism comprises a switch unit, an industrial personal computer and a PLC (programmable logic controller) which are in communication connection with the switch unit, and an algorithm server which is in communication connection with the switch unit; the code reading mechanism is in communication connection with the exchange unit, and the triggering mechanism is in communication connection with the PLC;

the intelligent recognition receiving device further comprises a display mechanism which is in communication connection with the processing control mechanism, wherein the display mechanism is a display screen, and the display screen is in communication connection with the industrial personal computer.

Compared with the prior art, the invention has the beneficial effects that:

the invention relates to an intelligent recognition receiving method and an intelligent recognition receiving device thereof, wherein the receiving method comprises the following steps: after a cargo pallet to be signed passes through a triggering mechanism, triggering a shooting part to shoot a container placed on the cargo pallet to be signed to obtain a pallet image, then transmitting the pallet image to a processing control mechanism and carrying out target detection on a plurality of labels in the pallet image to obtain corresponding coordinate positions of the labels in the pallet image, and obtaining the number of the labels contained in the pallet image; then, the processing control mechanism carries out character recognition screening on the character string information of the tag to obtain a plurality of corresponding first box number information; at the same time, triggering a code reading shooting piece to read the two-dimensional codes on the labels to obtain corresponding character information, and then transmitting the character information to a processing control mechanism for screening to obtain a plurality of corresponding second box number information and the number of the two-dimensional codes on the current cargo pallet; the method comprises the steps of comparing the number of the two-dimensional codes with the number of the labels to confirm the receiving mode of the current cargo carrying vehicle; if the number of the first case number information and the second case number information are equal, matching the first case number information and the second case number information: the box number information which can be corresponding is matched in a one-to-one correspondence manner, and other box number information which cannot be corresponding is matched randomly, so that the condition of goods accumulation caused by character string information identification and screening errors is avoided, and the goods receiving efficiency is improved; if the number of the two pieces of box number information is unequal, carrying out one-to-one correspondence matching on the box number information which can be corresponding, and if the other pieces of box number information cannot be corresponding, not carrying out matching, and marking the box number information which is not matched in a carrier cargo box image according to the corresponding coordinate position of the box number information for distinguishing; and then comparing with the current receiving plan, if the current container matching data is in the current receiving plan and the container matching data is in a matching completed state, completing container receiving of the current cargo pallet, otherwise, not receiving and marking the current cargo pallet. The intelligent recognition receiving device can be applied to intelligent recognition receiving methods for intelligent recognition receiving, timely and efficient receiving is achieved, goods accumulation is avoided, and receiving efficiency is improved.

Drawings

FIG. 1 is a flow chart of an intelligent identifying and receiving method of the present invention;

FIG. 2 is a schematic diagram of an intelligent identifying and receiving device according to the present invention;

FIG. 3 is a schematic diagram of the trigger mechanism and the guide mechanism of the present invention;

FIG. 4 is an enlarged schematic view of section I of FIG. 3;

FIG. 5 is a schematic diagram of the structure of the code reading mechanism and the light supplementing mechanism of the present invention;

FIG. 6 is a schematic diagram of a process control mechanism connection according to the present invention;

FIG. 7 is a graph showing the change of the accuracy and recall of an intelligent recognition receiving method according to the present invention;

FIG. 8 is a P-R curve of an intelligent recognition receiving method according to the present invention.

The graphic indicia are illustrated as follows:

the device comprises a 1-code reading mechanism, a 11-portal frame, a 111-first connecting rod, a 112-first mounting seat, a 12-code reading shooting piece, a 13-shooting piece, a 2-triggering mechanism, a 21-first laser sensor, a 22-second laser sensor, a 23-third laser sensor, a 3-processing control mechanism, a 31-PLC (programmable logic controller), a 32-industrial personal computer, a 33-switch, a 34-server, a 4-guiding mechanism, a 41-first guide rail, a 411-first guiding structure, a 42-second guide rail, a 421-second guiding structure, a 43-third guide rail, a 5-light supplementing mechanism, a 51-second connecting rod, a 511-second mounting seat, a 52-light supplementing lamp and a 6-display mechanism.

Detailed Description

The invention is further described below in connection with the following detailed description. Wherein the drawings are for illustrative purposes only and are shown in schematic, non-physical, and not intended to be limiting of the present patent; for the purpose of better illustrating embodiments of the invention, certain elements of the drawings may be omitted, enlarged or reduced and do not represent the size of the actual product; it will be appreciated by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.

The same or similar reference numbers in the drawings of embodiments of the invention correspond to the same or similar components; in the description of the present invention, it should be understood that, if there is an azimuth or positional relationship indicated by terms such as "upper", "lower", "left", "right", etc., based on the azimuth or positional relationship shown in the drawings, it is only for convenience of describing the present invention and simplifying the description, but it is not indicated or implied that the apparatus or element referred to must have a specific azimuth, be constructed and operated in a specific azimuth, and thus terms describing the positional relationship in the drawings are merely illustrative and should not be construed as limitations of the present patent, and specific meanings of the terms described above may be understood by those skilled in the art according to specific circumstances.

Example 1

Fig. 1 shows a first embodiment of an intelligent recognition receiving method of the present invention, which includes the following steps:

s1, after a cargo pallet to be signed passes through a triggering mechanism 2, triggering a shooting piece 13 to shoot containers placed on the cargo pallet and obtain a pallet image of the pallet, wherein the pallet image of the pallet comprises all containers on the current cargo pallet, and each container is stuck with a label which comprises two-dimensional code information and character string information;

s2, transmitting the pallet truck container image to a processing control mechanism 3, and carrying out target detection on a plurality of labels in the pallet truck container image to obtain corresponding coordinate positions of the labels in the pallet truck container image and obtain the number of the labels contained in the pallet truck container image; then, the processing control mechanism 3 carries out OCR (optical character recognition) on the character string information of the tag to obtain a plurality of corresponding first box number information;

s3, after the goods trolley to be signed passes through the triggering mechanism 2, the code reading shooting piece 12 is triggered to read the two-dimensional code on the label, and corresponding character information is obtained; the character information is transmitted to the processing control mechanism 3 for screening to obtain a plurality of corresponding second box number information and the number of the two-dimensional codes on the current cargo pallet;

s6, carrying out matching processing on the first box number information and the second box number information: carrying out one-to-one correspondence matching on box number information which can be corresponding, carrying out no matching on the rest box number information which cannot be corresponding, marking the box number information which cannot be matched in a cargo box image of the cargo box according to the corresponding coordinate position of the box number information, then summarizing and packaging cargo box matching data of the current cargo box, and then executing step S7;

s7, comparing the container matching data which is completed and packaged with the current day receiving plan, if the current container matching data is in the current day receiving plan and the container matching data is in a matching completed state, completing container receiving of the current cargo pallet, otherwise, not receiving and marking the current cargo pallet.

The invention comprises an intelligent recognition receiving method, which can be applied to the inspection and acceptance of goods with large goods quantity such as parts, and the like, and comprises the steps of triggering a shooting part 13 to shoot a container placed on a goods pallet to be signed to obtain a pallet image of the pallet, then transmitting the pallet image of the pallet to a processing control mechanism 3, and carrying out target detection on a plurality of labels in the pallet image of the pallet to obtain corresponding coordinate positions of the labels in the pallet image of the pallet, and obtaining the number of the labels contained in the pallet image of the pallet; then, the processing control mechanism 3 performs character recognition screening on the character string information of the tag to obtain a plurality of corresponding first box number information; at the same time, the code reading shooting piece 12 is triggered to read the two-dimensional code on the label to obtain corresponding character information, and then the character information is transmitted to the processing control mechanism 3 for screening to obtain a plurality of corresponding second box number information and the number of the two-dimensional codes on the current cargo pallet; the triggering mechanism 2 is arranged, so that unnecessary false touch to the shooting piece 13 can be avoided, unnecessary code scanning operation of code reading shooting and 12 can be avoided, the image processing quantity and code scanning processing are reduced, and the processing efficiency is improved; then, by comparing the number of the labels with the number of the two-dimension codes on the labels, confirming whether the labels on each container currently have abnormal conditions, if not, namely, when the number of the labels is equal, the two-dimension codes in each label are not missing, and at the moment, matching the first box number information acquired through the two-dimension codes with the second box number information acquired through the character strings: the two box number information can be in one-to-one correspondence, one-to-one correspondence matching is carried out, the other box number information is subjected to random matching, and the random matching arrangement can avoid the situation that the corresponding box number information cannot be accurately identified due to character string identification errors or incomplete character strings caused by improper image acquisition, so that goods cannot be received, can avoid excessive goods accumulation, and improves goods receiving efficiency; if the two-dimension code is abnormal, namely when the two numbers are unequal, the two-dimension code is missing in each label, when the first box number information is matched with the second box number information, the part of the first box number information and the second box number information, which are inconsistent, is not matched, the box number information, which is not matched, is marked in the cargo box image of the carrier according to the corresponding coordinate position of the box number information, and the two-dimension code is used for distinguishing other cargo receiving conditions so that operators can check later, and then the cargo receiving is not carried out.

Example 2

The present embodiment is similar to embodiment 1 except that the trigger mechanism 2 in this embodiment includes a first trigger assembly, a second trigger assembly, and a third trigger assembly that are disposed in this order along the direction of conveyance of the cargo pallet. The step S1 specifically comprises the following steps: the processing control mechanism 3 receives the first in-place signal, the second in-place signal and the third in-place signal of the cargo pallet detected by the first triggering component, the second triggering component and the third triggering component and judges whether the receiving sequence of the first in-place signal and the second in-place signal is the sequence of the first in-place signal, the second in-place signal and the third in-place signal; if so, triggering the code reading shooting piece and the shooting piece to shoot, and obtaining a cargo box image of the carrier; otherwise, if the reverse direction exists, photographing is not performed, and the transportation of the current cargo pallet is readjusted.

The step S3 specifically comprises the following steps: the processing control mechanism 3 receives the first in-place signal, the second in-place signal and the third in-place signal of the cargo truck detected by the first triggering component, the second triggering component and the third triggering component, judges whether the receiving sequence of the first in-place signal and the second in-place signal is reverse, if not, triggers the code reading shooting component 12 to read the code, otherwise does not read the code.

It should be noted that, steps S1 and S3 are not sequential.

In this embodiment, the step S2 specifically includes the following steps: the process control means 3 includes a target detection model and a character recognition model; wherein: transmitting the pallet truck container image to a target detection model for target detection, specifically, the target detection model uses YOLOV5 as a basic technical frame, extracting corresponding characteristics of labels in a neural network by using a multi-layer network, classifying by scoring, positioning coordinate positions corresponding to each label in the pallet truck container image, and obtaining the number of labels contained in the pallet truck container image; and the character recognition model is a CRNN character recognition model, text regions in the labels are extracted through det text detection, character string information in the text regions is recognized through rec text recognition, characters conforming to a number/number form are screened out, and a plurality of corresponding first box number information is obtained. The second case number information obtained by screening by the processing control mechanism 3 after the code reading shooting part 12 reads the two-dimensional code information on the label is also character information in a form of "number/number".

It should be noted that, the target detection model adopts a large number of pictures shot on site as a training set, contains at least ten thousands of labels in total, trains 100 epochs, and finally obtains the model with the accuracy rate reaching 99.989% and the recall rate reaching 99.977%. The YOLOV5 of the invention has the advantage that in the aspect of data enhancement, a new Mosaic method is added for enriching the data set. In the aspect of the framework, the Pytorch framework can be used for conveniently training the data set, and the Pytorch framework is easier to put into production, so that the model has the characteristics of quick training, good effect, easiness in deployment and the like. In order to meet the required training project of the invention, the actual training of the epoch can reach the peak value after 100 rounds of accuracy, so the epoch is set to be 100-150. To increase the training speed, the batch-size is set to 16, and the screen size of the image is set to 1280, so that the training accuracy is not affected.

Also, the confidence threshold is set to 0.5 when the target detection model predicts, and some false detection results can be filtered out. And screening the detection result once by screening the detection frame with the confidence coefficient larger than a certain threshold value. The higher the confidence threshold is set, the more stringent the screening of the model detection results, and the more likely the label will be missed.

The steps of the object detection model in this embodiment include:

I. inputting the picture to be detected into a trained target detection model;

II, extracting the characteristics of the picture through a multilayer network in the target detection model;

thirdly, scoring the characteristics of the picture through a certain scoring mechanism, so as to classify, and positioning the coordinate position of the target through regression;

and IV, summarizing the classification result, the score, the coordinate position and the like, and outputting a model prediction result.

It should be further noted that, in this embodiment, the target detection model is accelerated through the TensorRT, and is optimized for the algorithm frame and the graphics card operation, so that acceleration of the deep neural network model can be achieved, the detection speed of the model is greatly improved, and meanwhile, no obvious loss of detection precision exists. The image reasoning time before and after the acceleration of the model is 135ms and 72ms respectively, so that the improvement is obvious.

In addition, the indexes for evaluating the target detection model include accuracy (Precision), recall (Recall), average accuracy (mAP) and detection speed. For a certain sample, four prediction scenarios are:

true Positive (TP) (correct tag detected), true Negative (TN) (no tag detected), false Positive (FP) (no tag detected, tag actually present), false Negative (FN) (tag detected, tag actually not present). The accuracy represents the correct proportion of the model to the detection of the predicted target in all the predicted frames, and reflects the correct detection capacity of the model:

the recall rate is the ratio of the model to the correct predicted frame in all real object frames, reflecting the model's ability to miss targets:

the change curves of each index in the training process of the target detection model are shown in fig. 7 and 8, and the accuracy and recall rate of the method can be obtained through fig. 7 and 8, so that the actual use requirements can be met.

Also, the present embodiment further includes a display mechanism 6, in which:

the step S5 specifically comprises the following steps: matching the first box number information with the second box number information: carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a first color wire frame in a display mechanism 6 according to the corresponding coordinate position of the label; and carrying out random matching on the rest of the labels which cannot be corresponded to, displaying a second color wire frame in the display mechanism 6 according to the corresponded coordinate positions of the labels, and then executing step S7.

Specifically, when the first box number information or the second box number information has repeated information and the repeated box number information part has a part which can be correspondingly matched, selecting one of the repeated box number information to be correspondingly matched with the other box number information which can be correspondingly matched for the part of the repeated information, and displaying a second color line frame in the display mechanism 6 according to the corresponding coordinate position of the label; the rest repeated box number information participates in random matching, and a second color wire frame is also displayed in the display mechanism 6 according to the corresponding coordinate position of the label; for the part with non-repeated information, the box number information which can be corresponding is matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism 6 according to the corresponding coordinate position of the label; and carrying out random matching on the rest of the labels which cannot be corresponded to, and displaying a second color wire frame in the display mechanism 6 according to the corresponded coordinate positions of the labels, and then executing step S7.

Specifically, when the first box number information has two identical character information in the form of 'number/number', and the second box number information has one character information identical to the repeated information, one of the two identical first box number information is arbitrarily selected to be matched with the second box number information correspondingly, and the other first box number information participates in random matching. The matching mode can avoid the situation that the corresponding box number information cannot be accurately identified due to the fact that the character string is incomplete due to the fact that the character string is identification errors or the image is obtained improperly, and further the situation that goods cannot be received occurs, and the matching mode is specifically expressed as follows: insufficient field illumination, light reflection, or shielding by foreign matters, or other emergency conditions, the character string cannot be completely acquired by the shooting piece 13, or the character string cannot be accurately identified due to insufficient definition of an image acquired by the shooting piece 13, and the like. Therefore, excessive goods can be prevented from being piled up in a random matching mode, and the goods receiving efficiency is improved. For randomly matched containers, the operator can reconfirm the corresponding container on the pallet truck in a subsequent operation because the color line frame of the container different from the normal shipment is identified at the display mechanism 6.

The step S6 specifically comprises the following steps:

s63, selecting one piece of repeated box number information to be correspondingly matched with the other piece of correspondingly matched box number information for the part with repeated information, and displaying a fourth color line frame in a display mechanism 6 according to the corresponding coordinate position of the label; the rest repeated box number information is not matched, and a fourth color wire frame is also displayed in the display mechanism 6 according to the corresponding coordinate position of the box number information which is not matched;

for the part with non-repeated information, the box number information which can be corresponding is matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism 6 according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the box number information which cannot be matched, and then the step S7 is executed;

Specifically, there is a case that: when the first box number information has two character information in the same number/number form and the second box number information has one character information identical to the repeated information, one of the two identical first box number information is selected to be matched with the second box number information at will, a fourth color line frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the label, the other first box number information is not matched, and the fourth color line frame is also displayed in the display mechanism 6 according to the corresponding coordinate position of the box number information which is not matched.

Specifically, there is another case that: when the second case number information has two character information in the same number/number form and the first case number information has one character information identical to the repeated information, one of the two identical second case number information is selected at will to be matched with the first case number information correspondingly, a fourth color line frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the label, the other second case number information is not matched, and the unmatched second case number information does not display the line frame in the display mechanism 6 because the unmatched second case number information does not correspond to the corresponding coordinate position of the label. However, since the other cases have a wire frame display, the tag that does not display a wire frame can be distinguished from the tag of the other cases.

S64, for the part where the information is repeated, displaying a third color wire frame in the display mechanism 6 according to the corresponding coordinate position; for the part with non-repeated information, the box number information which can be corresponding is matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism 6 according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the box number information which cannot be matched, and then the step S7 is executed;

specifically, there is a case that: when the first box number information has two character information in the same number/number form and the second box number information has no character information identical to the repeated information, a fourth color line frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the label for the two first box number information at the same time. For the part with non-repeated information, the first box number information and the second box number information which can be corresponding are matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism 6 according to the corresponding coordinate position of the label; and the rest of the first box number information which cannot be matched is not matched, a third color wire frame is displayed in the display mechanism 6 according to the corresponding coordinate position of the first box number information which cannot be matched, and the wire frame is not displayed when the second box number information which cannot be matched and does not correspond to the coordinate position is not matched.

Specifically, there is another case that: when the second box number information has two character information of the same "number/number" form and the first box number information has no character information identical to the above-described repeated information, the wire frame is not displayed in the display means 6 for both the second box number information. And for the part where the information is not repeated, the same processing as in the above case is performed.

It should be noted that, when the number of two-dimensional codes is not equal to the number of labels, the reason why the box number information which cannot be matched correspondingly is not randomly matched is as follows: when the two-dimension codes on the batch of cargo boxes are missing or the two-dimension codes are unequal in number due to other reasons, the batch of cargo boxes are provided with the procedures which further need to be processed in the next step, and the two-dimension codes need to be distinguished from the condition that the two-dimension codes are equal in number, so that matching processing is not performed.

S65, carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a first color wire frame in a display mechanism 6 according to the corresponding coordinate position of the label; and if the other box numbers cannot be matched correspondingly, the box number information which is not matched is displayed with a third color wire frame in the display mechanism 6 according to the corresponding coordinate position of the box number information, and then the step S7 is executed.

Example 3

Fig. 2 to 6 show a first embodiment of an intelligent recognition receiving device of the present invention, which includes a code reading mechanism 1, a triggering mechanism 2 and a processing control mechanism 3, wherein the code reading mechanism 1 includes a gantry 11 and a plurality of code reading shooting pieces 12 and shooting pieces 13 connected to the gantry 11, and shooting ends of the code reading shooting pieces 12 and the shooting pieces 13 are all used for facing a cargo carrying vehicle; the triggering mechanism 2 comprises a plurality of sensing assemblies arranged along the conveying direction of the cargo pallet; the code reading shooting piece 12, the shooting piece 13 and the sensing component are all in communication connection with the processing control mechanism 3.

When the goods-carrying vehicle to be signed passes through the plurality of sensing components of the triggering mechanism 2, the processing control mechanism 3 receives signals of the sensing components, then the code reading shooting piece 12 and the shooting piece 13 of the code reading mechanism 1 are controlled to shoot and read codes for a plurality of containers on the goods-carrying vehicle, and then the acquired images are transmitted to the processing control mechanism 3 for information analysis and processing.

As shown in fig. 5, a first connecting rod 111 is vertically disposed in the gantry 11, and the code reading shooting piece 12 and the shooting piece 13 are connected with the first connecting rod 111 through a first mounting seat 112. In this embodiment, the first connecting rod 111 may be used as a first connecting rail structure, the first mounting seat 112 may be used as a first sliding seat, the code reading shooting piece 12 and the shooting piece 13 are both slidably connected with the first connecting rail structure through the first sliding seat, and the height positions of the code reading shooting piece 12 and the shooting piece 13 can be adjusted through the sliding arrangement. A locking structure is further arranged between the first sliding seat and the first connecting guide rail structure, and the locking effect of the locking structure can be used for positioning the code reading shooting piece 12 and the shooting piece 13 on the first connecting rod 111. It should be noted that the locking structure may be any locking manner such as bolt locking and buckle locking, which is not limited herein.

In order to improve the photographing accuracy of the code reading photographing member 12, the code reading photographing member 12 is provided along the first connecting rod 111 with at least two. In this embodiment, the code reading shooting pieces 12 are code reading cameras, and three code reading cameras are equidistantly arranged along the first connecting rod 111; the photographing part 13 is an industrial camera, which is disposed on the first connection rod 111 at a height position of the code reading camera near the middle. It should be noted that, the position of the industrial camera near the middle and the equidistant setting of the three code reading cameras are only preferred, and the specific height positions of the industrial camera and each code reading camera can be adjusted according to the actual shooting field of view.

As shown in fig. 3, the intelligent recognition receiving device further includes a guiding mechanism 4, where the guiding mechanism 4 includes a first guide rail 41, a second guide rail 42 and a third guide rail 43 that are disposed along the conveying direction of the cargo pallet, and the three are parallel to each other, and the leading-in ends of the first guide rail 41, the second guide rail 42 and the third guide rail 43 are respectively provided with a first leading-in structure 411, a second leading-in structure 421 and a third leading-in structure. In this embodiment, the first introducing structure 411, the second introducing structure 421 and the third introducing structure are all in a wedge-like structure, one end of the wedge-like structure is a wide end surface, the wide end surface is fixedly connected with the end of each guide rail, the other end of the wedge-like structure is a narrow end surface, and the narrow end surface is used as the free end of each guide rail, so that the introduction of the cargo carrier can be facilitated. The narrow end face of the wedge-like structure means an end having a smaller cross-sectional area. In this embodiment, the cargo pallet may be a forklift, and correspondingly, the first guide rail 41, the second guide rail 42, and the third guide rail 43 may be ground ox guide rails, which may be fastened to the ground by bolts; the arrangement of the intervals between the first rail 41, the second rail 42 and the third rail 43 may be set according to the wheel positions of the actual forklift.

As shown in fig. 3 and 4, the trigger mechanism 2 includes a first laser sensor 21, a second laser sensor 22, and a third laser sensor 23 disposed along the guide direction of a second guide rail 42, and the first laser sensor 21, the second laser sensor 22, and the third laser sensor 23 are also disposed in this order along the guide direction of the third guide rail 43. When a pallet truck loaded with a cargo box enters the guide mechanism 4, the first, second and third laser sensors 21, 22 and 23 can be correspondingly triggered according to the current arrival position of the pallet truck. The position setting and the interval setting of each laser sensor in the embodiment can be adjusted according to the actual use situation.

As shown in fig. 2, two code reading mechanisms 1 are provided in this embodiment, and a guide mechanism 4 is located between the two code reading mechanisms 1. The arrangement of the two code reading mechanisms 1 can be used for carrying out code reading identification on the cargo information on two sides of the cargo pallet, so that the cargo receiving efficiency is improved.

As shown in fig. 6, the processing control mechanism 3 includes a switch unit 31, an industrial personal computer 32 electrically connected to the switch unit 31, a PLC controller 33, and an algorithm server 34 electrically connected to the switch unit 31; the code reading photographing element 12 and the photographing element 13 are electrically connected with the exchange unit 31, and the first laser sensor 21, the second laser sensor 22 and the third laser sensor 23 are electrically connected with the PLC controller 33.

As shown in fig. 2, the intelligent identifying and receiving device further includes a display mechanism 6 in communication connection with the processing control mechanism 3, specifically, the display mechanism 6 is a display screen, the display screen is electrically connected with the industrial personal computer 32, and the identifying result of the cargo box on the current cargo carrying vehicle can be displayed through the display screen.

Example 4

The embodiment is similar to embodiment 3, except that, as shown in fig. 5, the intelligent recognition receiving device in the embodiment further includes a light supplementing mechanism 5 communicatively connected to the processing control mechanism 3, and a light source emitting end of the light supplementing mechanism 5 is used for facing the cargo pallet. In this embodiment, the light supplementing mechanism 5 includes a second connecting rod 51 and a plurality of light supplementing lamps 52 movably connected to the second connecting rod 51, where the light supplementing lamps 52 are electrically connected to the switch unit 31 through a light source controller.

As shown in fig. 5, in this embodiment, the light compensating lamp 52 is movably connected with the second connecting rod 51 through the second mounting seat 511, and the height and the light compensating angle position of the light compensating lamp 52 can be adjusted through the arrangement of the movable connection. The second mounting seat 511 may be connected to the second connecting rod 51 by a locking structure, and the locking effect of the locking structure may be used to position the light compensating lamp 52 on the second connecting rod 51. Specifically, the locking structure may be a C-shaped hoop, where the opening of the hoop is locked by a bolt, so that the second mounting seat 511 can rotate with the rotation shaft of the second connecting rod 51.

The invention discloses an intelligent recognition receiving device which can be applied to an intelligent recognition receiving method, and specifically comprises the following steps:

s1, a cargo pallet to be signed enters a code reading mechanism 1 along a guide mechanism 4, after passing through a triggering mechanism 2, if a first laser sensor 21, a second laser sensor 22 and a third laser sensor 23 are triggered in sequence, a PLC (programmable logic controller) 33 and an exchange unit 31 industrial personal computer 32 trigger a light supplementing lamp 52 of a light supplementing mechanism 5 to supplement light, and a PLC 33 and an exchange unit 31 trigger a shooting piece 13 to shoot a cargo container placed on the cargo pallet, so that a pallet container image of the pallet is obtained; the carrier cargo box image comprises labels stuck on each cargo box, and the labels comprise two-dimensional code information and character string information; at the same time, the code reading shooting piece 12 is triggered to read the two-dimensional code on the label to obtain corresponding character information;

s2, transmitting the pallet truck container image to an algorithm server 34 through the exchange unit 31, wherein a target detection model and a character recognition model are stored in the algorithm server 34, and performing target detection on a plurality of labels in the pallet truck container image through the target detection model to obtain corresponding coordinate positions of the labels in the pallet truck container image and obtain the number of the labels contained in the pallet truck container image; further, character recognition screening is carried out on the character string information of the tag through a character recognition model, so that a plurality of corresponding first box number information are obtained;

S3, transmitting the corresponding character information obtained by reading the two-dimension codes to an algorithm server 34 through a switch unit 31, and screening the character information by the algorithm server 34 to obtain a plurality of corresponding second box number information and the number of the two-dimension codes on the current cargo pallet;

s5, carrying out matching processing on the first box number information and the second box number information: carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a green wire frame in a display screen according to the corresponding coordinate position of the label; the rest can not be matched randomly, an orange wire frame is displayed in a display screen according to the corresponding coordinate position of the label, then container matching data of the current cargo pallet is summarized and packed, and then step S7 is executed;

s6, carrying out matching processing on the first box number information and the second box number information: carrying out one-to-one correspondence matching on the box number information which can correspond to the box number information, and displaying a green wire frame in a display screen according to the corresponding coordinate position of the label; the other box number information which cannot be matched is not matched, corresponding color line frame display is carried out on the box number information which cannot be matched in the cargo box image of the carrier according to the corresponding coordinate position of the box number information, and line frame display is not carried out on the first box number information which does not correspond to the corresponding coordinate position of the label; then summarizing and packaging container matching data of the current cargo pallet, and then executing step S7;

It is to be understood that the above examples of the present invention are provided by way of illustration only and not by way of limitation of the embodiments of the present invention. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the invention are desired to be protected by the following claims.

Claims

1. The intelligent recognition receiving method is characterized by comprising the following steps of:

s1, after a cargo pallet to be signed passes through a triggering mechanism (2), triggering a shooting piece (13) to shoot cargo boxes placed on the cargo pallet and obtain a pallet box image of the pallet, wherein the pallet box image of the pallet comprises all cargo boxes on the current cargo pallet, and each cargo box is stuck with a label, and the label comprises two-dimensional code information and character string information;

S2, transmitting the trolley container image to a processing control mechanism (3), and carrying out target detection on a plurality of labels in the trolley container image to obtain corresponding coordinate positions of the labels in the trolley container image and obtain the number of the labels contained in the trolley container image; then, the processing control mechanism (3) performs character recognition screening on the character string information of the tag to obtain a plurality of corresponding first box number information;

s3, after the goods carrying vehicle to be signed passes through the triggering mechanism (2), the code reading shooting piece (12) is triggered to read the two-dimensional code on the label, and corresponding character information is obtained; the character information is transmitted to the processing control mechanism (3) for screening to obtain a plurality of corresponding second box number information and the number of the two-dimensional codes on the current cargo pallet;

2. The intelligent recognition receiving method according to claim 1, wherein the triggering mechanism (2) comprises a first triggering component, a second triggering component and a third triggering component which are sequentially arranged along the conveying direction of the cargo pallet;

the step S1 specifically comprises the following steps: the processing control mechanism (3) receives the first in-place signal, the second in-place signal and the third in-place signal of the cargo pallet detected by the first triggering component, the second triggering component and the third triggering component, judges whether the receiving sequence of the first in-place signal and the second in-place signal is reverse, if not, triggers the shooting component (13) to shoot and obtains the cargo container image of the pallet, otherwise does not shoot;

The step S3 specifically includes the following steps: the processing control mechanism (3) receives first in-place signals, second in-place signals and third in-place signals of the cargo pallet detected by the first triggering component, the second triggering component and the third triggering component, judges whether the receiving order of the first in-place signals and the second in-place signals is reverse, if not, triggers the code reading shooting component (12) to read codes, otherwise does not read codes;

and the steps S1 and S3 are not sequential.

3. The intelligent recognition receiving method according to claim 1, wherein the processing control mechanism (3) comprises a target detection model and a character recognition model;

the step S2 specifically includes the following steps: transmitting the pallet truck container image to the target detection model for target detection, obtaining coordinate positions corresponding to all labels in the pallet truck container image, and obtaining the number of labels contained in the pallet truck container image; and then, using the character recognition model to recognize character string information in the tag, and screening out characters conforming to the form of numbers to obtain a plurality of corresponding second box number information.

4. The intelligent recognition receiving method according to claim 3, wherein the object detection model uses YOLOV5 as a basic technical framework, extracts corresponding features of the tag in a neural network by using a multi-layer network, classifies the features by scoring, and then locates coordinate positions of the tag; the character recognition model is a CRNN character recognition model, text regions in the labels are extracted through det text detection, and character string information in the text regions is recognized and screened through rec text recognition.

5. The intelligent recognition receiving method according to claim 1, further comprising a display mechanism (6), wherein:

the step S5 specifically includes the following steps: matching the first box number information with the second box number information: carrying out one-to-one correspondence matching on the box number information which can be corresponding to the box number information, and displaying a first color wire frame in the display mechanism (6) according to the corresponding coordinate position of the label; the rest can not be matched randomly, a second color wire frame is displayed in the display mechanism (6) according to the corresponding coordinate position of the label, and then the goods receiving is completed;

the step S6 specifically includes the following steps:

s63, selecting one piece of repeated box number information to be correspondingly matched with the other piece of correspondingly matched box number information for the part with repeated information, and displaying a fourth color line frame in the display mechanism (6) according to the corresponding coordinate position of the label; the rest repeated box number information is not matched, and a fourth color wire frame is also displayed in the display mechanism (6) according to the corresponding coordinate position of the box number information which is not matched;

For the part with non-repeated information, the box number information which can be corresponding is matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism (6) according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in a display mechanism (6) according to the corresponding coordinate position of the box number information which cannot be matched, and then step S7 is executed;

s64, for the part with repeated information, displaying a third color wire frame in the display mechanism (6) according to the corresponding coordinate position; for the part with non-repeated information, the box number information which can be corresponding is matched in a one-to-one correspondence manner, and a first color wire frame is displayed in a display mechanism (6) according to the corresponding coordinate position of the label; the rest of the box number information which cannot be matched is not matched, a third color wire frame is displayed in a display mechanism (6) according to the corresponding coordinate position of the box number information which cannot be matched, and then step S7 is executed;

s65, carrying out one-to-one correspondence matching on the box number information which can be corresponding to the box number information, and displaying a first color wire frame in a display mechanism (6) according to the corresponding coordinate position of the label; and if the other box numbers cannot be matched correspondingly, the box number information which is not matched is displayed with a third color wire frame in a display mechanism (6) according to the corresponding coordinate position of the box number information, and then step S7 is executed.

6. The intelligent recognition receiving device is characterized by comprising a code reading mechanism (1), a triggering mechanism (2) and a processing control mechanism (3), wherein the code reading mechanism (1) comprises a portal (11) and a plurality of code reading shooting pieces (12) and shooting pieces (13) which are connected to the portal (11), and shooting ends of the code reading shooting pieces (12) and the shooting pieces (13) are both used for facing a cargo carrying vehicle; the triggering mechanism (2) comprises a plurality of sensing assemblies arranged along the conveying direction of the cargo pallet; the code reading shooting piece (12), the shooting piece (13) and the sensing assembly are all in communication connection with the processing control mechanism (3).

7. The intelligent recognition receiving device according to claim 6, wherein a first connecting rod (111) is arranged in the portal frame (11), the code reading shooting piece (12) and the shooting piece (13) are connected to the first connecting rod (111) in a sliding mode, and locking structures are arranged between the code reading shooting piece (12) and the first connecting rod (111) and between the shooting piece (13) and the first connecting rod (111); the first connecting rod (111) is vertically arranged, the code reading shooting pieces (12) are arranged at least two along the first connecting rod (111), the code reading shooting pieces (12) are code reading cameras, and the shooting pieces (13) are industrial cameras.

8. The intelligent recognition receiving device according to claim 6, further comprising a guide mechanism (4), wherein the guide mechanism (4) comprises a plurality of guide rails arranged along the conveying direction of the cargo pallet, and the leading-in ends of the guide rails are provided with leading-in structures; the plurality of sensing assemblies are sequentially arranged on the guide rail along the extending direction of the guide rail; the number of the code reading mechanisms (1) is two, and the guide rail is positioned between the two code reading mechanisms (1).

9. The intelligent recognition receiving device according to claim 6, further comprising a light supplementing mechanism (5) in communication connection with the processing control mechanism (3), wherein a light source emitting end of the light supplementing mechanism (5) is used for facing the cargo carrying vehicle; the light supplementing mechanism (5) comprises a second connecting rod (51) and a plurality of light supplementing lamps (52) movably connected to the second connecting rod (51), and the light supplementing lamps (52) are in communication connection with the processing control mechanism (3) through a light source controller.

10. Intelligent recognition receiving device according to any of claims 6 to 9, characterized in that the processing control mechanism (3) comprises a group of switches (31) and an industrial personal computer (32) and a PLC controller (33) which are in communication connection with the group of switches (31), and further comprises an algorithm server (34) in communication connection with the group of switches (31); the code reading mechanism (1) is in communication connection with the exchange unit (31), and the triggering mechanism (2) is in communication connection with the PLC (33);

The intelligent recognition receiving device further comprises a display mechanism (6) which is in communication connection with the processing control mechanism (3), wherein the display mechanism (6) is a display screen, and the display screen is in communication connection with the industrial personal computer (32).