Abnormality tracing method and system

Abstract

The invention relates to the technical field of storage resource management, and particularly discloses a storage abnormity tracing method and system, and the method comprises the steps: obtaining the cargo information at a warehousing end in real time, and building a cargo label and a cargo feature based on the cargo information; creating a cache database with cargo labels as names, acquiring warehouse videos in real time based on a camera system, and positioning cargo areas in the warehouse videos based on cargo features; identifying a cargo area, determining cargo variation and workers in each time period, and inputting the cargo variation and workers into a cache database; receiving a traceability request which is input by a worker and contains a cargo label and a variable quantity, and performing traceability in the cache database; according to the method, the cargo state is acquired in real time through the camera in the warehouse, the cargo change amount and nearby workers are determined according to the cargo state, the nearby workers are regarded as cargo change reasons and set as responsible personnel, and when source tracing is needed, direct query can be carried out.

Description

Technical Field

[0001] This application is a divisional application of an invention application with an application date of April 19, 2024, a Chinese application number of 202410474005.1, and an invention name of "A Method and System for Traceability of Warehouse Abnormalities". Background Art

[0002] "Warehousing" refers to the process of storing and managing goods, which usually occurs in warehouses or logistics centers. In the business field, warehousing is an important part of supply chain management, involving activities such as receiving, storing, picking, packaging, and distributing goods.

[0003] Most existing warehousing processes adopt a responsible model, where each batch of goods is managed by full-time staff. Once a problem occurs, it is handled by the full-time staff; this method is essentially a dedicated person for a dedicated position, relying too much on the ability of the staff. When the quantity of goods is large, the physical and mental pressure of the staff is very high. How to provide a more relaxed warehousing state attribution solution and reduce the pressure of the staff is the technical problem that the technical solution of the present invention wants to solve. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for traceability of warehouse abnormalities to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A method for traceability of warehouse abnormalities, the method includes:

[0007] Real-time obtain goods information at the inbound end, and create a goods label and goods characteristics based on the goods information; the goods information includes goods transfer information and goods physical information; the goods transfer information is used to represent the source and destination of the goods, and the goods physical information includes goods weight and goods volume; the goods characteristics are the characteristics of the goods in the image.

[0008] Create a cache database named after the goods label, obtain the warehouse video in real time based on the camera system, and locate the goods area in the warehouse video based on the goods characteristics.

[0009] Identify the goods area, determine the goods change amount and staff at each time period, and input them into the cache database.

[0010] Receive a traceability request containing the goods label and change amount input by the staff, and perform traceability in the cache database.

[0011] As a further solution of the present invention: the step of real-time obtaining goods information at the inbound end and creating a goods label and goods characteristics based on the goods information includes:

[0012] At the warehousing end, it receives in real time the name of the goods, the goods transfer information, and the physical information of the goods uploaded by the staff, and determines the goods number according to the name of the goods, the upload time, and the goods transfer information as the goods label;

[0013] Query the standard image according to the name of the goods, and calculate the environmental filter according to the goods image and the standard image;

[0014] Obtain the goods image at the warehousing end, correct the goods image according to the environmental filter, and locate the goods area in the corrected goods image;

[0015] Extract the goods features in the located goods area and connect them to the goods label.

[0016] As a further solution of the present invention: The steps of creating a cache database named after the goods label, obtaining the warehouse video in real time based on the camera system, and locating the goods area in the warehouse video based on the goods features include:

[0017] Create a cache database named after the goods label and query the goods features corresponding to the goods label;

[0018] Obtain the warehouse video in real time by the camera system and convert the warehouse video into image frames;

[0019] Traverse each image based on the goods features to obtain the matching goods position;

[0020] Taking the goods position as the center, expand the goods area;

[0021] Among them, the process of expanding the goods area with the goods position as the center is:

[0022] Calculate the center of the goods position, query the nearest obstacle in each direction according to the preset angular step, and synchronously calculate the nearest distance; determine the extension length in this direction according to the nearest distance; the extension length is:

[0023] In the formula, D θ is the extension length in the θ direction, and d θ is the nearest distance in the θ direction.

[0024] As a further solution of the present invention: The steps of traversing each image based on the goods features to obtain the matching goods position include:

[0025] Convert the goods features to the frequency domain and select the pixel value range in the frequency domain;

[0026] Split the goods features into sub-features of different sizes based on the pixel value range; the size of the sub-feature is represented by an information ratio, which is a percentage and is used to characterize the ratio of the data volume of the sub-feature to the data volume of the goods feature;

[0027] Select the smallest sub-feature and traverse each image. When the match is successful, read the sub-features in ascending order of size and perform a traversal match. When the largest sub-feature matches successfully, use the position where the match is successful as the goods position;

[0028] Among them, the traversal order of the image is the rotation order, including the clockwise order and the counterclockwise order; when a goods position is obtained, use the goods position as the traversal starting point for the next image;

[0029] The pixel value range is (μ - iσ, μ + iσ); in the formula, μ is the mean value of each pixel value in the frequency domain, σ is the standard deviation of each pixel value, and i is a value determined by the size of the sub-feature, and i is proportional to the size of the sub-feature.

[0030] As a further solution of the present invention: the steps of identifying the goods area, determining the goods change amount and the staff at each time period, and inputting into the cache database include:

[0031] Identify the goods area and determine the goods position and its volume at each moment;

[0032] Obtain the goods weight with a timestamp according to the weighing points pre-installed in the warehouse;

[0033] Statistical the volume of goods at each time, fit a volume change curve, and correct the volume change curve according to the goods weight with a timestamp;

[0034] Take the derivative of the corrected volume change curve. When the derivative value reaches a preset value, select a preset-length independent variable interval centered on the independent variable corresponding to the derivative value;

[0035] Read the goods area corresponding to the independent variable interval, identify the goods area, locate the staff, and determine the responsibility level according to the distance between the staff and the goods position;

[0036] Statistical the staff with responsibility levels, the independent variable intervals, and the goods change amounts in the independent variable intervals, sort the statistical data based on the left endpoint of the independent variable interval, and input it into the cache database.

[0037] As a further solution of the present invention: the steps of receiving a traceability request containing a goods label and a change amount input by a staff member and performing traceability in the cache database include:

[0038] Receive a traceability request containing a goods label and a change amount input by a staff member;

[0039] Based on the goods label positioning cache database, traverse the cache database based on the change amount, and match the independent variable interval and the staff members with responsibility levels.

[0040] The technical solution of the present invention also provides a warehousing anomaly traceability system, and the system includes:

[0041] An information filing module, configured to obtain goods information in real time at the warehousing end, and create a goods label and goods features based on the goods information; the goods information includes goods circulation information and goods physical information; the goods circulation information is used to characterize the source and destination of the goods, and the goods physical information includes the goods weight and the goods volume; the goods features are the features of the goods in the image;

[0042] A video acquisition and recognition module, configured to create a cache database named after the goods label, obtain the warehouse video in real time based on the camera system, and locate the goods area in the warehouse video based on the goods features;

[0043] A data backup module, configured to identify the goods area, determine the change amount of the goods and the staff members at each time period, and input them into the cache database;

[0044] A data traceability module, configured to receive a traceability request containing the goods label and the change amount input by the staff member, and perform traceability in the cache database.

[0045] As a further solution of the present invention: the information filing module includes:

[0046] A label generation unit, configured to receive the goods name, goods circulation information and goods physical information uploaded by the staff member in real time at the warehousing end, and determine the goods number according to the goods name, upload time and goods circulation information as the goods label;

[0047] A filter generation unit, configured to query the standard image according to the goods name, and calculate the environmental filter according to the goods image and the standard image;

[0048] A region positioning unit, configured to obtain the goods image at the warehousing end, correct the goods image according to the environmental filter, and locate the goods area in the corrected goods image;

[0049] A feature extraction unit, configured to extract the goods features in the located goods area and connect them to the goods label.

[0050] As a further solution of the present invention: the video acquisition and recognition module includes:

[0051] A goods feature query unit, configured to create a cache database named after the goods label and query the goods features corresponding to the goods label;

[0052] A video conversion unit, configured to obtain a warehouse video in real time by a camera system and convert the warehouse video into image frames;

[0053] A position matching unit, configured to traverse each image based on the goods features to obtain the matching positions of the goods;

[0054] A region expansion unit, configured to expand a goods region with the goods position as the center;

[0055] Wherein, the process of expanding a goods region with the goods position as the center is as follows:

[0056] Calculate the center of the goods position, query the nearest obstacle in each direction according to a preset angular step size, and synchronously calculate the nearest distance; determine the extension length in this direction according to the nearest distance; the extension length is:

[0057] In the formula, D θ is the extension length in the θ direction, and d θ is the nearest distance in the θ direction.

[0058] As a further solution of the present invention: the position matching unit includes:

[0059] A range selection sub-unit, configured to convert the goods features into the frequency domain and select a pixel value range in the frequency domain;

[0060] A feature splitting sub-unit, configured to split the goods features into sub-features of different sizes based on the pixel value range; the size of the sub-feature is represented by an information ratio, and the information ratio is a percentage, used to characterize the ratio of the data volume of the sub-feature to the data volume of the goods features;

[0061] A matching execution sub-unit, configured to select the smallest sub-feature, traverse each image, when the matching is successful, read the sub-features in ascending order of size for traversal and matching, and when the largest sub-feature matches successfully, use the successfully matched position as the goods position;

[0062] Wherein, the traversal order of the images is a rotation order, including a clockwise order and a counterclockwise order; when a goods position is obtained, use the goods position as the traversal starting point of the next image;

[0063] The pixel value range is (μ - iσ, μ + iσ); in the formula, μ is the mean value of each pixel value in the frequency domain, σ is the standard deviation of each pixel value, and i is a value determined by the size of the sub-feature, and i is directly proportional to the size of the sub-feature.

[0064] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention obtains the status of goods in real time through cameras in the warehouse, determines the change amount of goods and the nearby staff according to the status of goods, regards the nearby staff as the cause of the change of goods, sets them as responsible personnel, and can directly query when traceability is required. The present invention has strong real-time performance, does not set full-time personnel, and there are no over-stressed staff, and is extremely suitable for existing intelligent warehouses. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0066] Figure 1 The flowchart of a method for tracing the source of warehouse anomalies is shown.

[0067] Figure 2 The block diagram of the composition structure of a system for tracing the source of warehouse anomalies is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the following further details the present invention with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0069] Figure 1 The flowchart of a method for tracing the source of warehouse anomalies is shown. In an embodiment of the present invention, a method for tracing the source of warehouse anomalies is provided, and the method includes:

[0070] Step S100: Obtain the goods information in real time at the inbound end, and create a goods label and goods characteristics based on the goods information; the goods information includes the goods transfer information and the goods physical information; the goods transfer information is used to represent the source and destination of the goods, and the goods physical information includes the goods weight and the goods volume; the goods characteristics are the characteristics of the goods in the image;

[0071] Step S200: Create a cache database with the goods label as the name, obtain the warehouse video in real time based on the camera system, and locate the goods area in the warehouse video based on the goods characteristics;

[0072] Step S300: Identify the goods area, determine the change amount of goods and the staff at each time period, and input them into the cache database;

[0073] Step S400: Receive the traceability request containing the goods label and the change amount input by the staff, and perform traceability in the cache database.

[0074] The present invention obtains goods information at the warehousing end. The goods information includes goods circulation information and goods physical information. The goods circulation information is not essential information. It is mainly used to determine the characteristics of the goods, indicating from which location the goods are transported to which location and which transfer stations they pass through on the way. The goods physical information is the weight and volume of the goods. By analyzing the goods information, the unique label of the goods and the goods characteristics can be determined. Among them, the goods circulation information is used to determine the goods label, and the goods physical information is used to determine the goods characteristics and correct the goods label. Correcting the goods label by the goods physical information can ensure the uniqueness of the goods label.

[0075] Then, for each batch of goods entering the warehouse, a cache database named after the goods label is established to store the identification data of the corresponding goods. The warehouse video is obtained by the camera system in the warehouse. Based on the goods characteristics, the goods area can be located in the warehouse video in real time to determine the location where this batch of goods appears in the warehouse at each moment.

[0076] By identifying each goods area, the quantity of goods and the nearby staff at each moment can be determined. The quantity of goods at each moment is counted and the change situation is calculated. At this time, it can be considered that the change situation is affected by the corresponding staff, thereby establishing a responsibility relationship.

[0077] Finally, when tracing is required, the staff inputs a tracing request, locates the corresponding cache database, and reads the change situation at each moment and the corresponding responsible personnel. This is the meaning of tracing in this application.

[0078] As a preferred embodiment of the technical solution of the present invention, step S100 is defined. The steps of obtaining goods information in real time at the warehousing end and creating goods labels and goods characteristics based on the goods information include:

[0079] At the warehousing end, the goods name, goods circulation information and goods physical information uploaded by the staff are received in real time, and the goods number is determined according to the goods name, upload time and goods circulation information as the goods label.

[0080] Query the standard image according to the goods name, and calculate the environmental filter according to the goods image and the standard image.

[0081] Obtain the goods image at the warehousing end, correct the goods image according to the environmental filter, and locate the goods area in the corrected goods image.

[0082] Extract the goods characteristics in the located goods area and connect them to the goods label.

[0083] The warehousing end is generally a computer device with information receiving function. When staff members warehousing goods, the computer device receives the goods name and goods transfer information input by the staff members, which is used to generate goods labels. To ensure the uniqueness of the generated goods labels, this application also records the input time of the staff members, connects these information, and then combines with a conventional conversion model to convert them into digital form, and then the goods labels can be obtained.

[0084] Furthermore, query the standard image of the goods by the goods name. The standard image is an image taken under normal light conditions of the goods. Comparing it with the goods image can determine the influence of the environment on the goods, which is called environmental filter. Correcting the goods image by the environmental filter can convert the goods image into an image under normal light conditions. At this time, identifying the goods image can obtain a more accurate result.

[0085] Identifying the corrected image can locate the goods area and extract the goods features.

[0086] Furthermore, the steps of creating a cache database named after the goods label, and based on the camera system to obtain the warehouse video in real time and locate the goods area in the warehouse video based on the goods features include:

[0087] Create a cache database named after the goods label, and query the goods features corresponding to the goods label;

[0088] Obtain the warehouse video in real time by the camera system, and convert the warehouse video into image frames;

[0089] Traverse each image based on the goods features to obtain the matching goods position;

[0090] Expand the goods area with the goods position as the center;

[0091] Among them, the process of expanding the goods area with the goods position as the center is:

[0092] Calculate the center of the goods position, query the nearest obstacle in each direction according to the preset angle step, and synchronously calculate the nearest distance; determine the extension length in this direction according to the nearest distance; the extension length is:

[0093] D θ =(1-e -dθ )d θ ; In the formula, D θ is the extension length in the θ direction, d θ is the nearest distance in the θ direction.

[0094] For each batch of goods, a cache database is created with the name of the goods label, the goods features corresponding to the goods label are queried, and the goods features are traversed and located in the warehouse video to query the positions of the goods at different times; centering on the goods position, an extended area is determined. The extended area is a range, and the personnel within the range can be regarded as the responsible personnel of the goods at different times.

[0095] Among them, regarding the goods features, the existing feature extraction scheme can be adopted for the goods features. The simplest scheme is to only denoise the goods area located in the goods image, and then perform contour recognition, and directly use the recognized contour as the goods features. In the subsequent matching process, it is only necessary that the similarity between the two parties to be matched reaches a small value.

[0096] As a preferred embodiment of the technical solution of the present invention, the step of traversing each image based on the goods features to obtain the matching goods position includes:

[0097] Convert the goods features to the frequency domain and select the pixel value range in the frequency domain;

[0098] Based on the pixel value range, split the goods features into sub-features of different sizes; the size of the sub-feature is represented by the information ratio, and the information ratio is a percentage, which is used to characterize the ratio of the data volume of the sub-feature to the data volume of the goods features;

[0099] Select the smallest sub-feature, traverse each image, when the matching is successful, read the sub-features in ascending order of size for traversal and matching. When the largest sub-feature matches successfully, the matching position is used as the goods position;

[0100] Among them, the traversal order of the images is the rotation order, including the clockwise order and the counterclockwise order; when a goods position is obtained, the goods position is used as the traversal starting point of the next image;

[0101] The pixel value range is (μ - iσ, μ + iσ); in the formula, μ is the mean value of each pixel value in the frequency domain, σ is the standard deviation of each pixel value, and i is a value determined by the size of the sub-feature, and i is directly proportional to the size of the sub-feature.

[0102] In an example of the technical solution of the present invention, the application process of the goods features is limited. The goods features are converted to the frequency domain, and then the pixel value range is selected in the frequency domain. Pixel points are extracted from the image features according to the selected pixel value range to obtain sub-features with different amounts of information. The sub-features are used to traverse the image. When the matching is successful, check whether the larger features can match. If the largest sub-feature can also match, it is considered that the goods have been matched, and the matched goods position can be recorded.

[0103] The above processing method can greatly improve the matching efficiency. The logic for improving the matching efficiency is that when the smaller sub-features do not match, it is defaulted that the corresponding positions no longer match, and there is no need to read the larger features for matching.

[0104] Specifically, regarding the traversal order, when there is a matching position, in the next image, the present application starts from this matching position and traverses clockwise or counterclockwise around it. This is also a way to improve the matching efficiency because the speed during the goods transportation process is stable and does not mutate, and the matching positions in the front and back images are similar.

[0105] As a preferred embodiment of the technical solution of the present invention, the steps of identifying the goods area, determining the goods change amount and the staff at each time period, and inputting into the cache database include:

[0106] Identifying the goods area to determine the goods position and its volume at each moment;

[0107] Obtaining the timestamp-containing goods weight according to the weighing points pre-installed in the warehouse;

[0108] Statistical volume of goods at each time, fitting out the volume change curve, and correcting the volume change curve according to the timestamp-containing goods weight;

[0109] Taking the derivative of the corrected volume change curve. When the derivative value reaches a preset value, a preset-length independent variable interval is selected with the independent variable corresponding to the derivative value as the center;

[0110] Reading the goods area corresponding to the independent variable interval, identifying the goods area, positioning the staff, and determining the responsibility level according to the distance between the staff and the goods position;

[0111] Statistical staff with responsibility levels, independent variable intervals, and the goods change amount in the independent variable intervals, sorting the statistical data based on the left endpoint of the independent variable interval, and inputting into the cache database.

[0112] The above content specifically defines the identification process of the goods area. First, identify the goods area to determine the goods position and its volume at each moment. Among them, the goods position has been known during the matching process, and the goods volume only needs to identify the size of the goods in the image, which is very easy under the background of the existing technology. By counting the goods volume at each moment, a volume change curve can be fitted. During the goods transportation process, there may be some weighing nodes in the warehouse. Detect the goods weight at the weighing nodes, and the fitted volume change curve can be corrected by the goods weight. Since the density of the goods is known, the correctness of the identified volume can be determined from the goods weight and the density of the goods, and then some accurate coordinates can be inserted into the volume change curve to improve the authenticity of the volume change curve.

[0113] Derive the volume change curve to calculate the volume change of the goods at each moment. The volume change reflects the increase or decrease of the goods. When the derivative is large, it indicates that the increase or decrease of the goods is more obvious. At this time, taking the independent variable (corresponding to time) with the larger derivative value as the center, obtain a time interval, read the goods area within the time interval, identify the goods area, locate the staff, and they can be used as the responsible persons for the current increase or decrease situation. According to the distance between the staff and the goods, determine the responsibility levels of different staff. Generally, the closer the distance, the higher the responsibility level.

[0114] Finally, count the staff with responsibility levels, the independent variable interval, and the goods change amount in the independent variable interval, and input them into the cache database for subsequent problem tracing. It should be noted that the goods change amount is determined by the goods volume change amount within the independent variable interval. The goods volume change amount can be directly used as the goods change amount, or multiplied by the density to use the goods weight change amount as the goods change amount.

[0115] As a preferred embodiment of the technical solution of the present invention, the steps of receiving the tracing request containing the goods label and the change amount input by the staff and performing tracing in the cache database include:

[0116] Receive the tracing request containing the goods label and the change amount input by the staff;

[0117] Locate the cache database based on the goods label, traverse the cache database based on the change amount, and match the independent variable interval and the staff with responsibility levels.

[0118] The above content is the final application process. The staff who needs to trace inputs the tracing request and uploads the goods label and the change amount. The goods label is used to locate the cache database, and the change amount is used to query the staff and their responsibility levels in the cache database.

[0119] Figure 2The block diagram of the composition structure of a warehousing anomaly traceability system is shown. In an embodiment of the present invention, a warehousing anomaly traceability system is provided. The system 10 includes:

[0120] An information filing module 11, configured to obtain goods information in real time at the inbound end, and create a goods label and goods features based on the goods information; the goods information includes goods circulation information and goods physical information; the goods circulation information is used to characterize the source and destination of the goods, and the goods physical information includes the goods weight and the goods volume; the goods features are the features of the goods in the image.

[0121] A video acquisition and recognition module 12, configured to create a cache database named after the goods label, obtain the warehouse video in real time based on the camera system, and locate the goods area in the warehouse video based on the goods features.

[0122] A data backup module 13, configured to identify the goods area, determine the goods change amount and the staff at each time period, and input them into the cache database.

[0123] A data traceability module 14, configured to receive a traceability request containing the goods label and the change amount input by the staff, and perform traceability in the cache database.

[0124] Furthermore, the information filing module 11 includes:

[0125] A label generation unit, configured to receive the goods name, goods circulation information and goods physical information uploaded by the staff in real time at the inbound end, and determine the goods number as the goods label according to the goods name, upload time and goods circulation information.

[0126] A filter generation unit, configured to query the standard image according to the goods name, and calculate the environmental filter according to the goods image and the standard image.

[0127] A region positioning unit, configured to obtain the goods image at the inbound end, correct the goods image according to the environmental filter, and locate the goods area in the corrected goods image.

[0128] A feature extraction unit, configured to extract the goods features in the located goods area and connect them to the goods label.

[0129] Specifically, the video acquisition and recognition module 12 includes:

[0130] A goods feature query unit, configured to create a cache database named after the goods label, and query the goods features corresponding to the goods label.

[0131] A video conversion unit, configured to obtain the warehouse video in real time by the camera system, and convert the warehouse video into image frames.

[0132] A position matching unit, configured to traverse each image based on the cargo features to obtain the matching cargo position;

[0133] A region expansion unit, configured to expand a cargo region with the cargo position as the center;

[0134] Among them, the process of expanding a cargo region with the cargo position as the center is as follows:

[0135] Calculate the center of the cargo position, query the nearest obstacle in each direction according to a preset angular step size, and synchronously calculate the nearest distance; determine the extension length in this direction according to the nearest distance; the extension length is:

[0136] In the formula, D θ is the extension length in the θ direction, and d θ is the nearest distance in the θ direction.

[0137] Regarding the position matching unit in the above content, the position matching unit includes:

[0138] A range selection sub-unit, configured to convert the cargo features to the frequency domain and select a pixel value range in the frequency domain;

[0139] A feature splitting sub-unit, configured to split the cargo features into sub-features of different sizes based on the pixel value range; the size of the sub-feature is represented by an information ratio, and the information ratio is a percentage, which is used to characterize the ratio of the data volume of the sub-feature to the data volume of the cargo features;

[0140] A matching execution sub-unit, configured to select the smallest sub-feature, traverse each image, when the matching is successful, read the sub-features in ascending order of size for traversal matching, and when the largest sub-feature matches successfully, use the successfully matched position as the cargo position;

[0141] Among them, the traversal order of the images is the rotation order, including the clockwise order and the counterclockwise order; when a cargo position is obtained, use the cargo position as the traversal starting point of the next image;

[0142] The pixel value range is (μ - iσ, μ + iσ); in the formula, μ is the mean value of each pixel value in the frequency domain, σ is the standard deviation of each pixel value, and i is a value determined by the size of the sub-feature, and i is directly proportional to the size of the sub-feature.

[0143] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be included in the patent protection scope of the present invention by the same token.

Claims

1. An abnormality tracing method, characterized in that: The method comprises: Obtain cargo information in real time at the warehousing end, and establish cargo labels and cargo features based on the cargo information; the cargo information includes cargo flow information and cargo physical information; the cargo flow information is used to characterize the source and destination of the cargo, and the cargo physical information includes cargo weight and cargo volume; the cargo features are the features of the cargo in the image; Create a cache database named after the cargo tag, obtain warehouse videos in real time based on the camera system, and locate the cargo area in the warehouse video based on the cargo features; Identify the cargo area, determine the cargo changes and staff in each period, and enter them into the cache database; Receive the traceability request containing the goods label and the change amount input by the staff, and perform traceability in the cache database; Among them, the process of identifying the cargo area and determining the cargo change volume and staff in each time period includes: identifying the cargo area, determining the cargo location and cargo volume at each time, counting the cargo volume at each time, fitting the volume change curve, and taking the derivative of the volume change curve to calculate the volume change of the cargo at each time, and determining the time interval based on the volume change; reading the cargo area within the time interval, identifying the cargo area, locating the staff, and simultaneously determining the responsibility level of the staff; the responsibility level is inversely proportional to the distance; In the process of fitting the volume change curve, the weight of the goods at the weighing node is obtained, and the fitted volume change curve is corrected.

2. The abnormality tracing method according to claim 1, characterized in that: The steps of obtaining cargo information in real time at the warehousing end and establishing cargo labels and cargo features based on the cargo information include: Receive the name of goods, goods flow information and goods physical information uploaded by the staff in real time at the warehouse entry end, and determine the goods number according to the name of goods, upload time and goods flow information as the goods label; Query the standard image according to the name of the goods, and calculate the environment filter based on the goods image and the standard image; Acquire a cargo image at the warehousing end, correct the cargo image according to the environmental filter, and locate the cargo area in the corrected cargo image; The cargo features are extracted from the located cargo area and connected to the cargo labels.

3. The abnormality tracing method according to claim 2 is characterized in that: The steps of creating a cache database named after the cargo label, acquiring warehouse video in real time based on the camera system, and locating the cargo area in the warehouse video based on the cargo features include: Create a cache database named after the cargo label and query the cargo features corresponding to the cargo label; The camera system acquires warehouse videos in real time and converts the warehouse videos into image frames; Traverse each converted image frame based on the cargo features to obtain the matching cargo location; Taking the cargo location as the center, expand the cargo area; The process of expanding the cargo area with the cargo location as the center is as follows: Calculate the center of the cargo position, query the nearest obstacle in each direction according to the preset angle step, and simultaneously calculate the closest distance; determine the extended length in that direction according to the closest distance; the extended length is: Where D θ is the extension length in the θ direction, d θ is the shortest distance in the θ direction.

4. The abnormality tracing method according to claim 3 is characterized in that: The step of traversing each converted image frame based on the cargo feature to obtain the matching cargo location comprises: Convert the cargo features to the frequency domain and select the pixel value range in the frequency domain; The cargo feature is split into sub-features of different sizes based on the pixel value range; the size of the sub-feature is represented by an information ratio, which is a percentage and is used to represent the ratio of the data volume of the sub-feature to the data volume of the cargo feature; Select the smallest sub-feature and traverse each image. When the match is successful, read the sub-features in ascending order of size and traverse the matching. When the largest sub-feature is matched successfully, the matching position is used as the cargo position. The traversal order of the image is a rotation order, including clockwise and counterclockwise order; when a cargo position is obtained, the cargo position is used as the traversal starting point of the next image; The pixel value range is (μ-iσ,μ+iσ); where μ is the mean of the pixel values ​​in the frequency domain, σ is the standard deviation of the pixel values, and i is a value determined by the sub-feature size, and i is proportional to the sub-feature size.

5. The abnormality tracing method according to claim 1, characterized in that: The steps of identifying the cargo area, determining the cargo change amount and staff in each time period, and inputting the data into the cache database include: Identify the cargo area and determine the cargo location and volume at each moment; Obtain the weight of goods with time stamp based on the weighing points pre-installed in the warehouse; Count the cargo volume at each time, fit the volume change curve, and correct the volume change curve according to the cargo weight with timestamp; The modified volume change curve is derived, and when the derivative value reaches a preset value, an independent variable interval of a preset length is selected with the independent variable corresponding to the derivative value as the center; Read the cargo area corresponding to the independent variable interval, identify the cargo area, locate the staff, and determine the responsibility level based on the distance between the staff and the cargo location; Statistics include staff members with responsibility levels, independent variable intervals, and changes in goods in the independent variable intervals. The statistical data are sorted based on the left endpoint of the independent variable interval and input into the cache database.

6. The abnormality tracing method according to claim 5, characterized in that: The step of receiving a traceability request containing a cargo label and a change amount input by a staff member and performing traceability in a cache database includes: Receive the traceability request containing the goods label and the change quantity input by the staff; The cache database is located based on the cargo label, and the cache database is traversed based on the change amount to match the independent variable interval and the staff with responsibility levels.

7. An abnormality tracing system, characterized in that: The system comprises: An information filing module is used to obtain cargo information in real time at the warehousing end, and to establish cargo labels and cargo features based on the cargo information; the cargo information includes cargo circulation information and cargo physical information; the cargo circulation information is used to characterize the source and destination of the cargo, and the cargo physical information includes cargo weight and cargo volume; the cargo features are the features of the cargo in the image; The video acquisition and recognition module is used to create a cache database named after the cargo label, acquire warehouse videos in real time based on the camera system, and locate the cargo area in the warehouse video based on the cargo features; The data backup module is used to identify the cargo area, determine the cargo changes and staff in each period, and input them into the cache database; The data traceability module is used to receive traceability requests containing cargo labels and change quantities input by staff, and perform traceability in the cache database.

8. The abnormality tracing system according to claim 7, characterized in that: The information filing module includes: The label generation unit is used to receive the name of the goods, the flow information of the goods and the physical information of the goods uploaded by the staff in real time at the warehouse entry end, and determine the goods number as the goods label according to the name of the goods, the upload time and the flow information of the goods; A filter generation unit, used for querying a standard image according to a product name, and calculating an environment filter according to the product image and the standard image; A region positioning unit, used to obtain a cargo image at the warehousing end, correct the cargo image according to the environmental filter, and locate the cargo region in the corrected cargo image; The feature extraction unit is used to extract cargo features in the located cargo area and is connected to the cargo tag.

9. The abnormality tracing system according to claim 8, characterized in that: The video acquisition and recognition module includes: A cargo feature query unit, used to create a cache database named after the cargo label, and query the cargo features corresponding to the cargo label; A video conversion unit, used for acquiring warehouse videos in real time through a camera system, and converting the warehouse videos into image frames; A position matching unit, used for traversing each converted image frame based on the cargo feature to obtain a matching cargo position; The area expansion unit is used to expand the cargo area with the cargo location as the center; The process of expanding the cargo area with the cargo location as the center is as follows: Calculate the center of the cargo position, query the nearest obstacle in each direction according to the preset angle step, and simultaneously calculate the closest distance; determine the extended length in that direction according to the closest distance; the extended length is: Where D θ is the extension length in the θ direction, d θ is the shortest distance in the θ direction.

10. The abnormality tracing system according to claim 9, characterized in that: The position matching unit comprises: The range selection subunit is used to convert the cargo features into the frequency domain and select the pixel value range in the frequency domain; A feature splitting subunit, used for splitting the cargo feature into sub-features of different sizes based on the pixel value range; the size of the sub-feature is represented by an information ratio, which is a percentage and is used to represent the ratio of the data volume of the sub-feature to the data volume of the cargo feature; The matching execution subunit is used to select the smallest sub-feature, traverse each image, and when the match is successful, read the sub-features in increasing order of size, perform traversal matching, and when the largest sub-feature is matched successfully, the matching position is used as the cargo position; The traversal order of the image is a rotation order, including clockwise and counterclockwise order; when a cargo position is obtained, the cargo position is used as the traversal starting point of the next image; The pixel value range is (μ-iσ,μ+iσ); where μ is the mean of the pixel values ​​in the frequency domain, σ is the standard deviation of the pixel values, and i is a value determined by the sub-feature size, and i is proportional to the sub-feature size.

Images