Logistics sorting field anti-loss system and method based on video monitoring

Abstract

The present invention discloses a system and method for preventing loss of a logistics sorting yard based on video surveillance, and relates to the technical field of logistics sorting. If there is an abnormal situation in the sorting yard, the corresponding area is regarded as an abnormal area and an alarm instruction is issued, and the corresponding abnormal picture is pushed to the outside; after collecting the cargo loss data in the abnormal area, the loss degree of the cargo in the loss period is constructed by the loss data. If the loss degree exceeds expectations, the label information of the cargo is matched with the video image at each verification node, and the matching degree in the abnormal area is generated by the cargo matching data. If the matching degree does not exceed the matching threshold, the corresponding abnormal area is regarded as a divergence area; the operating coefficient of the monitoring device is constructed by the operating status data, and if the operating coefficient of the monitoring device is lower than the operating threshold, a maintenance instruction is issued to the outside. By maintaining the monitoring device, the operating status of the monitoring device is improved, the monitoring effect is improved, and the risk of cargo loss is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics sorting, and in particular to a logistics sorting field anti-loss system and method based on video monitoring. Background Art

[0002] Logistics sorting refers to the process of classifying, sorting and separating goods during logistics transportation so that they can be accurately distributed to different destinations. The main steps of logistics sorting include: receiving, sorting, marking, sorting, packaging, etc. The logistics sorting system can be divided into two sorting forms: "people to goods" and "goods to people". Among them, the automatic sorting system has been widely used in the logistics industry with its high efficiency and accuracy. In addition, there are some new sorting technologies, such as paperless selection technology, RF barcode recognition terminal technology, etc., which also play an important role in improving the efficiency and accuracy of logistics sorting. Logistics sorting systems are widely used in e-commerce logistics, express sorting, postal express, production and manufacturing and other fields. In these fields, the logistics sorting system uses automation equipment, machine vision and other technologies to achieve fast and accurate sorting operations, improving overall operational efficiency and customer satisfaction.

[0003] The Chinese invention patent with the authorization announcement number CN107824463B discloses an automatic logistics sorting system and an automatic logistics sorting method, including: an automatic transport unit, which transports items to be sorted; a sorting area, in which the automatic transport unit operates to sort the items to be sorted, the sorting area includes a drop zone with multiple sorting target objects, and the automatic transport unit performs the drop of the items to be sorted to the corresponding sorting target objects in the drop zone; and a control system, which controls the sorting of the items to be sorted and the movement of the automatic transport unit in the sorting area, and further includes: a setting unit, which is used to set a specific sorting target object for abnormal processing among multiple sorting target objects; a determination unit, which is used to determine whether the items to be sorted have an abnormality; and a control unit, which controls to transport the items to be sorted to the specific sorting target object in response to the abnormality of the items to be sorted. Through the automatic processing mechanism, efficient item sorting can be achieved.

[0004] Combined with the above application and the contents of the prior art:

[0005] When sorting goods in a sorting yard, due to the large amount of goods and the long sorting process, goods are often lost in the sorting yard, resulting in certain economic losses. In order to reduce the risk of goods lost in the sorting yard, the existing anti-loss method is usually to add label information to the goods, for example, attach QR codes and bar codes. After the goods enter different sorting stages and areas, the label information of the goods is read by a scanning device, and the area and location of the goods are determined based on the information reading process and results. Although this method can reduce the risk of goods lost, since the label information is relatively single in this verification method and cannot be mutually verified, it is difficult to accurately judge whether the goods have been lost, which leads to a high risk of goods lost.

[0006] To this end, the present invention provides a logistics sorting field anti-loss system and method based on video monitoring. Summary of the invention

[0007] 1. Technical issues to be solved

[0008] In view of the deficiencies of the prior art, the present invention provides a logistics sorting field anti-loss system and method based on video monitoring. After collecting the cargo loss data in the abnormal area, the loss degree of the cargo within the loss period is constructed from the loss data. If the loss degree exceeds expectations, the label information of the cargo is matched with the video image at each verification node, and the matching degree in the abnormal area is generated from the cargo matching data. If the matching degree does not exceed the matching threshold, the corresponding abnormal area is used as a divergence area; the operating coefficient of the monitoring device is constructed from the operating status data. If the operating coefficient of the monitoring device is lower than the operating threshold, a maintenance instruction is issued to the outside. By maintaining the monitoring equipment, the operating status of the monitoring equipment is improved, the monitoring effect is improved, and the risk of cargo loss is reduced. Thereby solving the technical problems raised in the background technology.

[0009] (II) Technical solution

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0011] A method for preventing loss in a logistics sorting yard based on video surveillance, including installing monitoring equipment in the sorting area and building a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, the monitoring blind spot is eliminated by adjusting the position and number of monitoring equipment;

[0012] The trained abnormality recognition model recognizes the real-time monitoring image data and obtains the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area is regarded as an abnormal area and an alarm instruction is issued, and the corresponding abnormal picture is pushed to the outside;

[0013] After collecting the cargo loss data in the abnormal area, the loss degree Sot of the cargo within the loss period is constructed based on the loss data. If the loss degree Sot exceeds expectations, a verification instruction is issued to the outside.

[0014] At each verification node, the label information of the goods is matched with the video image, and the matching degree Ps in the abnormal area is generated by the goods matching data. If the matching degree Ps does not exceed the matching threshold, the corresponding abnormal area is regarded as the divergence area; wherein, the matching degree Ps in the abnormal area is constructed by the goods matching data set, and after obtaining the matching rate G of the goods at each verification node, the following method is used:

[0015]

[0016] Among them, G i is the matching rate of the i-th verification node, G a is the target value of the matching rate, σ is the standard deviation of several matching rates; weight coefficient: 0≤S1≤1, 0≤S2≤1, and S1+S2=1;

[0017] The operating status data of the monitoring equipment in the divergence area is detected and acquired, and the operating coefficient Vt of the monitoring equipment is constructed based on the operating status data. If the operating coefficient Vt of the monitoring equipment is lower than the operating threshold, a maintenance instruction is issued to the outside.

[0018] Furthermore, an electronic map covering the sorting yard is constructed. After determining the key locations of the sorting yard and formulating a camera layout plan, monitoring equipment is installed at all key locations. Data collection is carried out in the sorting yard, and part of the collected data is extracted as sample data. The digital twin model of logistics sorting is obtained by training with the sample data.

[0019] Furthermore, after obtaining the location of each monitoring device in the sorting yard, the location of the monitoring device is marked on the logistics sorting digital twin model, and the logistics sorting digital twin model is used to simulate the image acquisition of the monitoring device to obtain the corresponding simulated monitoring image;

[0020] Several simulated monitoring images are spliced ​​together, and the spliced ​​images are compared with the area to be covered to determine whether there are any monitoring blind spots in the sorting yard. If there are any monitoring blind spots, they are marked on the electronic map and identified as monitoring blind spots, and additional monitoring equipment is installed in the monitoring blind spots.

[0021] Furthermore, after the monitoring equipment collects the image data of the goods in real time in the sorting area, the scanning equipment identifies the label information of the goods, and obtains the real-time location of the goods through image matching and analysis, and marks the real-time location of the goods in the logistics sorting digital twin model;

[0022] Use machine learning algorithm training to obtain an abnormality recognition model. After the collected monitoring image data is obtained by each monitoring device, it is input into the abnormality recognition model. The abnormality recognition model is used to identify the image data and obtain the corresponding recognition results to obtain abnormal situations in the sorting area.

[0023] Furthermore, after receiving the alarm instruction, the time when the alarm instruction is issued, the alarm area, the description of the abnormal situation and the corresponding abnormal picture in the sorting area are recorded; the alarm area and its adjacent areas are determined as abnormal areas, and the processing time, location, cargo information and management personnel information of the goods in the abnormal area are recorded; the above data are aggregated to construct a cargo sorting data set in the abnormal area.

[0024] Furthermore, if there is cargo loss in the abnormal area within the loss period, the cargo loss degree Sot within the loss period is obtained based on the cargo loss data analysis within the loss period, according to the following formula:

[0025]

[0026] Weight coefficient: 0≤β≤1, 0≤α≤1, and α+β=1, F ij is the shortest distance from the i-th lost area to the j-th lost area, F avg is the mean of the shortest distance, F iqr is the quartile of the shortest distance, P i is the price of the i-th lost item, P avg is the mean of the price; i = 1, 2, …, k, where k is the number of lost areas.

[0027] Furthermore, after receiving the verification instruction, the label information of the goods is matched with the video image at each verification node to confirm whether the goods information in the abnormal area is consistent during the sorting process. If inconsistent, the location and corresponding quantity information of the inconsistent goods are determined, and the goods matching data set is constructed after aggregation.

[0028] Furthermore, the divergence area is marked on the electronic map and used as a relative blind spot. After querying and obtaining the number of monitoring devices in the relative blind spot, the operating status of the corresponding monitoring device is detected, and the operating status data of the monitoring device is collected. The acquired equipment operating status data is summarized to construct an equipment operating status data set.

[0029] Furthermore, the operation coefficient Vt of the monitoring device is constructed from the device operation status data set, wherein the definition X of the image data obtained by the monitoring device is a And the data delay Y when the monitoring device sends image data to the outside a Perform linear normalization and map the corresponding data values ​​to the interval [0, 1], and then follow the following formula:

[0030]

[0031] Weight coefficient: 0≤χ≤1, 0≤ρ≤1, and χ+ρ=1.

[0032] The logistics sorting yard anti-loss system based on video surveillance includes: a blind spot inspection unit, which installs monitoring equipment in the sorting area and builds a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, the monitoring blind spot is eliminated by adjusting the position and number of monitoring equipment;

[0033] The abnormal behavior recognition unit uses the trained abnormal recognition model to recognize the real-time monitoring image data and obtain the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area will be regarded as an abnormal area and an alarm instruction will be issued, and the corresponding abnormal picture will be pushed to the outside;

[0034] The cargo loss analysis unit collects cargo loss data in the abnormal area, and constructs the cargo loss rate Sot within the loss period based on the loss data. If the loss rate Sot exceeds expectations, a verification instruction is issued to the outside.

[0035] The verification unit matches the label information of the goods with the video image at each verification node, and generates the matching degree Ps in the abnormal area from the goods matching data. If the matching degree Ps does not exceed the matching threshold, the corresponding abnormal area is regarded as the divergence area;

[0036] The equipment monitoring unit detects and obtains the operating status data of the monitoring equipment in the divergent area, and constructs the operating coefficient Vt of the monitoring equipment based on the operating status data. If the operating coefficient Vt of the monitoring equipment is lower than the operating threshold, a maintenance instruction is issued to the outside.

[0037] (III) Beneficial effects

[0038] The present invention provides a logistics sorting field anti-loss system and method based on video monitoring, which has the following beneficial effects:

[0039] 1. Determine whether there are blind spots in the sorting area. If there are blind spots, install monitoring equipment in the blind spots, or adjust the positions of monitoring equipment in adjacent areas to eliminate the blind spots in the sorting area. By achieving comprehensive monitoring, the risk of cargo loss can be reduced.

[0040] 2. Determine whether there is any abnormal situation in the sorting area. If there is any abnormal situation, issue an alarm command to the outside in time. After receiving the alarm command, the management personnel will extract the corresponding screen information. After obtaining the screen information, the management personnel can immediately process it to ensure the normal execution of the sorting process in the sorting yard and avoid the loss of goods.

[0041] 3. After identifying abnormalities in the surveillance video and screening out abnormal areas, the loss degree Sot is calculated and obtained. Based on the obtained loss degree Sot, a comprehensive assessment is made on the severity of cargo loss in the abnormal area. If the severity of the loss is severe, targeted processing is required.

[0042] 4. Confirm whether the cargo information during the sorting process is consistent with the information obtained by monitoring and identification. If the frequency of inconsistency is high, judge and evaluate the risk of cargo loss in the abnormal area based on the matching degree Ps. If the risk of cargo loss is high, further attention should be paid to the abnormal area to reduce the risk of cargo loss.

[0043] 5. Evaluate the operating status of the monitoring equipment based on the operating coefficient Vt. If the operating status of the monitoring equipment is poor, it will be difficult to achieve the desired monitoring effect, which will increase the risk of cargo loss. By maintaining the monitoring equipment, improving the operating status of the monitoring equipment, and improving the monitoring effect, the risk of cargo loss can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 This is a schematic diagram of the process of the method for preventing loss in a logistics sorting field of the present invention;

[0045] Figure 2 It is a schematic diagram of the structure of the anti-loss system of the logistics sorting yard of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] See also Figure 1 The present invention provides a method for preventing loss in a logistics sorting field based on video monitoring, comprising:

[0048] Step 1: Install monitoring equipment in the sorting area and build a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, eliminate the monitoring blind spot by adjusting the location and number of monitoring equipment;

[0049] The step 1 includes the following contents:

[0050] Step 101: construct an electronic map covering the sorting yard. After determining the key locations of the sorting yard, such as the entrance, exit, sorting area, and storage area, mark all the key locations on the electronic map, and after developing a camera layout plan, install monitoring equipment at all key locations, such as scanning equipment and high-definition cameras;

[0051] Step 102: collect data in the sorting yard, for example, data of various devices and their locations and aisle data in the sorting yard, and summarize the collected data. Extract part of the collected data as sample data, and use the sample data to train and test the initial model constructed by the machine learning algorithm to obtain a logistics sorting digital twin model.

[0052] Step 103: After obtaining the positions of each monitoring device in the sorting yard, mark the positions of the monitoring devices on the logistics sorting digital twin model, and use the logistics sorting digital twin model to simulate the image acquisition of the monitoring devices to obtain the corresponding simulated monitoring images;

[0053] Splice several simulated surveillance images, compare the spliced ​​images with the area to be covered, and determine whether there are surveillance blind spots in the sorting yard. If there are surveillance blind spots, mark them on the electronic map and identify them as surveillance blind spots, and install additional surveillance equipment in the surveillance blind spots;

[0054] When using, combine the contents in steps 101 to 103:

[0055] After entering the sorting state in the sorting yard, monitoring equipment is installed at various key locations in the sorting yard, and a logistics sorting digital twin model is constructed based on various data in the sorting yard. After the installation of the monitoring equipment is completed, the logistics sorting digital twin model verifies the installation position of the monitoring equipment to determine whether there are monitoring dead angles and blind spots in the sorting area. If there are monitoring blind spots, monitoring equipment can be installed in the monitoring blind spots, or the positions of monitoring equipment in adjacent areas can be adjusted to eliminate the monitoring blind spots in the sorting yard. When there are no monitoring blind spots for the goods in the sorting yard, comprehensive monitoring can be achieved to reduce the risk of loss of goods.

[0056] When sorting goods in a sorting yard, due to the large amount of goods and the long sorting process, goods are often lost in the sorting yard, resulting in certain economic losses. In order to reduce the risk of goods lost in the sorting yard, the existing anti-loss method is usually to add label information to the goods, for example, attach QR codes and bar codes. After the goods enter different sorting stages and areas, the label information of the goods is read by a scanning device, and the area and location of the goods are determined based on the information reading process and results. Although this method can reduce the risk of goods lost, since the label information is relatively single in this verification method and cannot be mutually verified, it is difficult to accurately judge whether the goods have been lost, which leads to a high risk of goods lost.

[0057] Step 2: The trained abnormality recognition model recognizes the real-time monitoring image data and obtains the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area is regarded as an abnormal area and an alarm instruction is issued, and the corresponding abnormal picture is pushed to the outside;

[0058] The step 2 includes the following contents:

[0059] Step 201: After the monitoring device collects the image data of the goods in real time in the sorting area, the scanning device identifies the label information of the goods, such as barcodes and QR codes, etc.; and obtains the real-time location of the goods through image matching and analysis, and marks the real-time location of the goods in the logistics sorting digital twin model; thereby, the location of the goods can be obtained in real time when the goods are in the sorting stage;

[0060] Step 202: Use machine learning algorithm training to obtain an abnormality recognition model. After each monitoring device obtains the collected monitoring image data, it is input into the abnormality recognition model, and the image data is recognized by the abnormality recognition model to obtain the corresponding recognition result; then, abnormal situations in the sorting area are recognized, analyzed and obtained, including abnormal behaviors of personnel and goods, such as no one in the sorting area, long retention time of goods, unauthorized entry, personnel activities at abnormal time or area, etc. If there is an abnormal situation in the sorting yard, an alarm instruction is issued to the outside, and the corresponding abnormal picture is pushed to the outside, such as to the management personnel;

[0061] When using, combine the contents in steps 201 and 202:

[0062] After the training and acquisition of the abnormal recognition model, the abnormal recognition model will recognize the real-time image data in each area of ​​the sorting yard, obtain the corresponding recognition results, and determine whether there is an abnormal situation in the sorting area. If there is an abnormal situation, an alarm instruction can be issued to the outside in time. After receiving the alarm instruction, the management personnel will extract the corresponding screen information. After obtaining the screen information, the management personnel can immediately process it to ensure the normal execution of the sorting process in the sorting yard and avoid the loss of goods.

[0063] Step 3: After collecting the cargo loss data in the abnormal area, the loss degree Sot of the cargo within the loss period is constructed based on the loss data. If the loss degree Sot exceeds expectations, a verification instruction is issued to the outside.

[0064] The step three includes the following contents:

[0065] Step 301: after receiving the alarm instruction, record the time when the alarm instruction is issued in the sorting area, the alarm area, the description of the abnormal situation and the corresponding abnormal screen; determine the alarm area and its adjacent areas as abnormal areas, and record the processing time, location, cargo information, management personnel information, etc. of the goods in the abnormal area; summarize the above data to construct a cargo sorting data set in the abnormal area, and update it in real time, so as to facilitate query and statistical analysis at any time;

[0066] Step 302: Set a loss period. If there is cargo loss in the abnormal area within the loss period, obtain the cargo loss degree Sot within the loss period based on the cargo loss data within the loss period. After obtaining the quantity, loss period, loss area and loss value of the lost cargo, the loss degree Sot is constructed according to the following formula under dimensionless conditions:

[0067]

[0068] Among them, F ij is the shortest distance from the i-th lost area to the j-th lost area, F avg is the mean of the shortest distance, F iqr is the quartile of the shortest distance, P i is the price of the i-th lost item, P avg is the mean of the price; i = 1, 2, ..., k, k is the number of lost areas; weight coefficient: 0 ≤ β ≤ 1, 0 ≤ α ≤ 1, and α + β = 1;

[0069] If the obtained loss degree Sot exceeds expectations, it means that more goods are lost in the abnormal area at the current stage. It is necessary to analyze and find the flow process and abnormal links of goods based on the goods sorting data in the goods sorting data set, and compare the location information of the lost goods. At this time, a verification instruction is issued to the outside;

[0070] When used, combine the contents in steps 301 and 302:

[0071] After identifying anomalies in the surveillance video and screening out the abnormal area, the data on cargo loss in the abnormal area in the recent period of time is collected. For example, after obtaining the approximate area, time period, quantity and value of the cargo loss, the loss degree Sot is calculated from these data. Based on the obtained loss degree Sot, a comprehensive assessment is made on the severity of cargo loss in the abnormal area. If the severity of the loss is severe, targeted processing is required.

[0072] Step 4: Match the label information of the goods with the video image at each verification node, and generate the matching degree Ps in the abnormal area from the goods matching data. If the matching degree Ps does not exceed the matching threshold, the corresponding abnormal area is regarded as the divergence area;

[0073] The step 4 includes the following contents:

[0074] Step 401: After receiving the verification instruction, a verification cycle including several verification nodes is set, and the label information of the goods is matched with the video image at each verification node to confirm whether the goods information in the abnormal area is consistent during the sorting process. If inconsistent, the location and corresponding quantity information of the inconsistent goods are determined, and the goods matching data set is constructed after aggregation;

[0075] Step 402: Construct the matching degree Ps in the abnormal area from the cargo matching data set, obtain the matching rate G of the cargo at each verification node, and then follow the following method:

[0076]

[0077] Among them, G i is the matching rate of the i-th verification node, G a is the target value of the matching rate, σ is the standard deviation of several matching rates; weight coefficient: 0≤S1≤1, 0≤S2≤1, and S1+S2=1; the weight coefficient can be obtained by referring to the hierarchical analysis method;

[0078] According to historical data and matching management expectations for cargo sorting, a matching threshold is pre-set; if the obtained matching degree Ps does not exceed the matching threshold, it means that there are many cargo errors in the corresponding abnormal area, and the cargo may be easily lost. At this time, the corresponding abnormal area is regarded as a divergence area;

[0079] When using, combine the contents in steps 401 and 402:

[0080] After the abnormal area is screened out in the sorting area, the various situations in the current abnormal area are identified to confirm whether the cargo information in the sorting process is consistent with the information obtained by monitoring and identification. If not, the cargo will be judged as lost or missed during further sorting and verification due to inconsistent cargo information, so there is a risk of loss. If the frequency of this inconsistency is high, the corresponding matching degree Ps is constructed based on the matching data, and the risk of cargo loss in the abnormal area is judged and evaluated based on the matching degree Ps. If the risk of cargo loss is high, further attention needs to be paid to the abnormal area to reduce the risk of cargo loss.

[0081] Step 5: Detect and obtain the operating status data of the monitoring equipment in the divergent area, and construct the operating coefficient Vt of the monitoring equipment based on the operating status data. If the operating coefficient Vt of the monitoring equipment is lower than the operating threshold, a maintenance instruction is issued to the outside;

[0082] The step five includes the following contents:

[0083] Step 501: Mark the divergence area on the electronic map and use it as a relative blind area. After querying and obtaining the number of monitoring devices in the relative blind area, detect the operating status of the corresponding monitoring device and collect monitoring device operating status data, including resolution, clarity, frame rate, lighting conditions and network connection, etc. The acquired device operating status data is aggregated to construct a device operating status data set;

[0084] Step 502: construct an operation coefficient Vt of the monitoring device from the device operation status data set, wherein the definition X of the image data obtained by the monitoring device is a And the data delay Y when the monitoring device sends image data to the outside a Perform linear normalization and map the corresponding data values ​​to the interval [0, 1], and then follow the following formula:

[0085]

[0086] Weight coefficient: 0≤χ≤1, 0≤ρ≤1, and χ+ρ=1;

[0087] Based on historical data and the expected operation management of the monitoring equipment, the operation threshold is pre-set; if the operation coefficient Vt is lower than the operation threshold, it can be said that its operation status is poor. When monitoring the flow of each cargo in the sorting yard, there may be screen loss or similar situations, which will increase the risk of cargo loss and require timely maintenance. At this time, a maintenance instruction can be issued to the outside;

[0088] After receiving the maintenance instructions, the monitoring equipment is quality inspected and equipment failures are eliminated in a timely manner, thereby avoiding monitoring blind spots and data loss, and preventing the loss of goods.

[0089] When using, combine the contents in steps 501 and 502:

[0090] After obtaining the area with a higher risk of cargo loss, the operating data of the corresponding monitoring equipment is obtained in the corresponding area, and several operating data are associated together to obtain the corresponding operating coefficient Vt. The operating status of the monitoring equipment is evaluated based on the operating coefficient Vt to determine whether its operating status is qualified and whether it can meet expectations. If the operating status of the monitoring equipment is poor or even fails, it is difficult to achieve the desired monitoring effect, which will increase the risk of cargo loss. As a corresponding processing method, the operating status of the monitoring equipment is improved and the monitoring effect is improved through maintenance of the monitoring equipment, so as to reduce the risk of cargo loss.

[0091] See also Figure 2 The present invention provides a logistics sorting field anti-loss system based on video monitoring, comprising:

[0092] Blind spot inspection unit: install monitoring equipment in the sorting area and build a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, adjust the location and number of monitoring equipment to eliminate the monitoring blind spot;

[0093] The abnormal behavior recognition unit uses the trained abnormal recognition model to recognize the real-time monitoring image data and obtain the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area will be regarded as an abnormal area and an alarm instruction will be issued, and the corresponding abnormal picture will be pushed to the outside;

[0094] The cargo loss analysis unit collects cargo loss data in the abnormal area, and constructs the cargo loss rate Sot within the loss period based on the loss data. If the loss rate Sot exceeds expectations, a verification instruction is issued to the outside.

[0095] The verification unit matches the label information of the goods with the video image at each verification node, and generates the matching degree Ps in the abnormal area from the goods matching data. If the matching degree Ps does not exceed the matching threshold, the corresponding abnormal area is regarded as the divergence area;

[0096] The equipment monitoring unit detects and obtains the operating status data of the monitoring equipment in the divergent area, and constructs the operating coefficient Vt of the monitoring equipment based on the operating status data. If the operating coefficient Vt of the monitoring equipment is lower than the operating threshold, a maintenance instruction is issued to the outside.

[0097] The Analytic Hierarchy Process (AHP) is a decision-making method that decomposes decision-related elements into levels such as goals, criteria, and plans, and conducts qualitative and quantitative analysis on this basis. It is particularly suitable for dealing with target systems with hierarchical and staggered evaluation indicators, and when the target value is difficult to describe quantitatively, the AHP is an effective decision-making tool.

[0098] The core of the hierarchical analysis method is to decompose the decision problem into multiple levels to form a hierarchical structure, which usually includes the target level, the criterion level, the sub-criterion level and the solution level. By solving the eigenvector of the judgment matrix, the priority weight of each element of each level to an element of the previous level is obtained, and the final weight of each alternative solution to the total goal is recursively merged by the weighted sum method, so as to find the optimal solution.

[0099] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0100] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0101] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0102] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0103] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0104] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0105] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0106] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for preventing loss in a logistics sorting field based on video surveillance, characterized in that: include, Install monitoring equipment in the sorting area and build a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, eliminate the monitoring blind spot by adjusting the location and number of monitoring equipment; The trained abnormality recognition model recognizes the real-time monitoring image data and obtains the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area is regarded as an abnormal area and an alarm instruction is issued, and the corresponding abnormal picture is pushed to the outside; After collecting the cargo loss data in the abnormal area, the loss rate of cargo within the loss period is constructed based on the loss data. If the degree of loss If the loss exceeds expectations, an external verification instruction will be issued; if there is cargo loss in the abnormal area within the loss period, the loss degree of the cargo will be obtained based on the cargo loss data analysis within the loss period. , according to the following formula: ； Weight coefficient: , ,and , For the Lost Area to The shortest distance of the lost area, is the mean of the shortest distances, is the quartile of the shortest distance, For the The price of the lost goods, is the mean of the prices; , is the number of lost areas; The cargo label information is matched with the video image at each verification node, and the cargo matching data generates the matching degree in the abnormal area. , if the matching degree If the matching threshold is not exceeded, the corresponding abnormal area is regarded as the divergence area; wherein the matching degree in the abnormal area is constructed by the cargo matching data set. , obtain the matching rate of goods at each verification node Then, follow the steps below: ； in, For the The matching rate of verification nodes, is the target value of the matching rate, is the standard deviation of several matching rates; weight coefficient: , ,and ; Detect and obtain the operating status data of the monitoring equipment in the divergent area, and construct the operating coefficient of the monitoring equipment based on the operating status data , if the operating coefficient of the monitoring equipment If the value is lower than the operation threshold, a maintenance instruction is issued to the outside, wherein the operation coefficient of the monitoring equipment is constructed by the equipment operation status data set. , where the image clarity and data delay Perform linear normalization and map the corresponding data values ​​to the interval Then follow the following formula: ； Weight coefficient: , ,and .

2. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 1 is characterized in that: Construct an electronic map covering the sorting yard. After determining the key locations of the sorting yard, formulate a camera layout plan and install monitoring equipment at all key locations. Collect data in the sorting yard, extract part of the collected data as sample data, and use the sample data to train and obtain the logistics sorting digital twin model.

3. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 2 is characterized in that: After obtaining the locations of each monitoring device in the sorting yard, mark the locations of the monitoring devices on the logistics sorting digital twin model, and use the logistics sorting digital twin model to simulate the image acquisition of the monitoring devices to obtain the corresponding simulated monitoring images; Several simulated monitoring images are spliced ​​together, and the spliced ​​images are compared with the area to be covered to determine whether there are any monitoring blind spots in the sorting yard. If there are any monitoring blind spots, they are marked on the electronic map and identified as monitoring blind spots, and additional monitoring equipment is installed in the monitoring blind spots.

4. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 3 is characterized in that: After the monitoring equipment collects the image data of the goods in real time in the sorting area, the scanning equipment identifies the label information of the goods, obtains the real-time location of the goods through image matching analysis, and marks the real-time location of the goods in the logistics sorting digital twin model; Use machine learning algorithm training to obtain an abnormality recognition model. After the collected monitoring image data is obtained by each monitoring device, it is input into the abnormality recognition model. The abnormality recognition model is used to identify the image data and obtain the corresponding recognition results to obtain abnormal situations in the sorting area.

5. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 4 is characterized in that: After receiving the alarm instruction, the time when the alarm instruction was issued, the alarm area, the description of the abnormal situation and the corresponding abnormal picture in the sorting area are recorded; the alarm area and its adjacent areas are determined as abnormal areas, and the processing time, location, cargo information and management personnel information of the goods in the abnormal area are recorded; the above data are aggregated to construct a cargo sorting data set in the abnormal area.

6. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 5 is characterized in that: After receiving the verification instruction, the label information of the goods is matched with the video image at each verification node to confirm whether the cargo information in the abnormal area is consistent during the sorting process. If inconsistent, the location and corresponding quantity information of the inconsistent goods are determined, and the cargo matching data set is constructed after aggregation.

7. The method for preventing loss in a logistics sorting field based on video surveillance according to claim 6 is characterized in that: The divergence area is marked on the electronic map and used as a relative blind spot. After querying and obtaining the number of monitoring devices in the relative blind spot, the operating status of the corresponding monitoring device is detected, and the operating status data of the monitoring device is collected. The acquired equipment operating status data is summarized to construct an equipment operating status data set.

8. A logistics sorting field anti-loss system based on video surveillance, applying the method described in any one of claims 1 to 7, characterized in that: include: Blind spot inspection unit: install monitoring equipment in the sorting area and build a digital twin model of logistics sorting in the sorting yard. If there is a monitoring blind spot in the sorting area, adjust the location and number of monitoring equipment to eliminate the monitoring blind spot; The abnormal behavior recognition unit uses the trained abnormal recognition model to recognize the real-time monitoring image data and obtain the corresponding recognition results. If there is an abnormal situation in the sorting yard, the corresponding area will be regarded as an abnormal area and an alarm instruction will be issued, and the corresponding abnormal picture will be pushed to the outside; The cargo loss analysis unit collects cargo loss data in the abnormal area and constructs the loss degree of cargo within the loss period based on the loss data. If the degree of loss If it exceeds expectations, an external verification instruction will be issued; The verification unit matches the cargo label information with the video image at each verification node, and generates the matching degree in the abnormal area based on the cargo matching data. , if the matching degree If the matching threshold is not exceeded, the corresponding abnormal area is regarded as the divergence area; The equipment monitoring unit detects and obtains the operating status data of the monitoring equipment in the divergent area, and constructs the operating coefficient of the monitoring equipment based on the operating status data , if the operating coefficient of the monitoring equipment If the voltage drops below the operating threshold, a maintenance instruction is issued to the outside.

Images