A method, device, equipment and medium for identifying bulk items

By using item identification cameras and pre-trained models in weighing equipment, identifying and classifying single bodies and categories of bulk items, the problem of weighing the same unit price for items is solved, and fast and accurate weighing is achieved, avoiding the Economic losses.

CN115077677BActive Publication Date: 2024-11-08YANTAI TRIAL RETAIL ENG CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210678279.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-06-13
Publication Date
2024-11-08
Estimated Expiration
2042-06-13

AI Technical Summary

Technical Problem

In supermarkets and other places, when customers purchase bulk items, due to unclear price tags and malicious mixing, items of different unit prices are weighed at the same unit price, causing economic losses, and increasing the workload for inspection and verification.

Method used

By configuring an item recognition camera and processor in a weighing device, using pre-trained item detection models and classification models, detecting and identifying the individuals and categories of weighing items, the density of weighing items is determined when the density is below the threshold. When the unit image recognition is performed to determine the item category to ensure accurate weighing.

Benefits of technology

It realizes accurate detection of various bulk items during the weighing process, avoids economic losses from customers and merchants, and improves weighing efficiency and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115077677B_ABST
    Figure CN115077677B_ABST
Patent Text Reader

Abstract

The present invention discloses a method, device, equipment and medium for identifying bulk items. The method comprises: if a weighing event of bulk items is detected, then obtaining the weighing item image collected by the weighing device, and performing a single-unit detection on the weighing item image based on a pre-trained item detection model to obtain a single-unit item detection result; determining the density of the weighing items according to the single-unit item detection result; if the density of the weighing items is lower than a preset density threshold, then obtaining a single-unit item image in the weighing item image, and determining the recognition result of the single-unit item image based on a pre-trained classification model. This technical solution can solve the problem of economic losses caused by weighing bulk items of different unit prices at the same unit price, and can accurately detect each bulk item during the weighing process, which is conducive to realizing fast weighing and avoiding economic losses for customers and merchants.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to a method, device, equipment and medium for identifying bulk items. Background Art

[0002] In supermarkets, community convenience stores and other places where goods are sold, bulk goods such as candies and biscuits account for a certain proportion. In order to save shopping bags, merchants usually allow customers to put bulk goods of the same price into the same shopping bag for weighing and settlement according to their personal preferences, such as putting candies of different flavors into one shopping bag and weighing them together.

[0003] However, when customers are shopping, it is easy for them to put items of different prices into the same shopping bag due to unclear price labels or malicious mixing. Therefore, when weighing and settling, one of the parties will suffer economic losses. At the same time, due to the wide variety of categories, the inspection and verification work also increases the workload for the weighers or cashiers. Summary of the invention

[0004] The present invention provides a bulk item identification method, device, equipment and medium to solve the problem of economic losses caused by weighing bulk items of different unit prices at the same unit price. Each bulk item can be accurately detected during the weighing process, which is conducive to rapid weighing and avoiding economic losses for customers and merchants.

[0005] According to one aspect of the present invention, there is provided a method for identifying bulk items, the method being performed by a weighing device, the method comprising:

[0006] If a weighing event of bulk items is detected, an image of the weighed item collected by the weighing device is obtained, and a single-item detection is performed on the image of the weighed item based on a pre-trained item detection model to obtain a single-item detection result;

[0007] Determining the density of weighed objects according to the detection results of the individual objects;

[0008] If the density of the weighed objects is lower than a preset density threshold, a single image of the object in the weighed object image is obtained, and a recognition result of the single image of the object is determined based on a pre-trained classification model.

[0009] According to another aspect of the present invention, there is provided a device for identifying bulk items, the device being configured in a weighing device, the device comprising:

[0010] A single-unit detection result generating module is used to obtain a weighing object image collected by the weighing device if a weighing event of a bulk object is detected, and to perform single-unit detection on the weighing object image based on a pre-trained object detection model to obtain a single-unit object detection result;

[0011] A density determination module, used to determine the density of weighed objects according to the detection results of the object monomers;

[0012] The recognition result determination module is used to obtain a single image of the weighed object in the weighed object image if the density of the weighed object is lower than a preset density threshold, and determine the recognition result of the single image of the object based on a pre-trained classification model.

[0013] According to another aspect of the present invention, there is provided a weighing device, characterized in that the device comprises: an object recognition camera, at least one processor and a memory;

[0014] The object recognition camera is communicatively connected to the processor, and the processor is communicatively connected to the memory;

[0015] Wherein, the object recognition camera is used to collect images of weighing objects;

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can implement the bulk article identification method described in any embodiment of the present invention when executing the computer program.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the bulk article identification method described in any embodiment of the present invention when executed.

[0018] The technical solution of the embodiment of the present invention, when a weighing event of bulk items is detected, obtains the weighing item image, and performs a single-unit detection on the weighing item image based on a pre-trained item detection model; then determines the density of the weighing items based on the single-unit detection result of the item; when the density of the weighing items is lower than the preset density threshold, obtains the single-unit image of the item in the weighing item image, and determines the recognition result of the single-unit image of the item based on a pre-trained classification model. This solution can solve the problem of economic losses caused by weighing bulk items of different unit prices at the same unit price, can accurately detect each bulk item during the weighing process, is conducive to rapid weighing, and avoids economic losses for customers and merchants.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of a bulk object identification method provided according to Embodiment 1 of the present invention;

[0022] Figure 2A is a flow chart of a bulk object identification method provided according to Embodiment 2 of the present invention;

[0023] Figure 2B is a schematic diagram of a method for determining the density of weighed objects provided in accordance with Embodiment 2 of the present invention;

[0024] Figure 3 is a schematic structural diagram of a bulk object identification device provided according to Embodiment 3 of the present invention;

[0025] Figure 4 It is a structural schematic diagram of a bulk article identification device for implementing the bulk article identification method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, device, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0028] Embodiment 1

[0029] Figure 1 A flowchart of a bulk object identification method is provided for the first embodiment of the present invention. This embodiment is applicable to the identification of bulk objects. The method can be performed by a bulk object identification device. The device can be implemented in the form of hardware and / or software. The device can be configured in a bulk object identification device. Figure 1 As shown, the method includes:

[0030] S110: If a weighing event of bulk items is detected, an image of the weighed item collected by the weighing device is obtained, and a single-item detection is performed on the image of the weighed item based on a pre-trained item detection model to obtain a single-item detection result.

[0031] This solution can be executed by a weighing device, which may include an object recognition camera for collecting images of the weighing objects. When bulk objects appear in the target area captured by the object recognition camera, it can be determined that a weighing event of the bulk objects has been detected. The bulk objects can be identified by the object recognition camera detecting at least two object monomers appearing in the target area. The weighing device can also set the application scope of the weighing device by setting up a store. For example, weighing device A is deployed in a bulk snack food store, and weighing device A is only used to weigh bulk snack foods. As long as weighing device A detects a change in weighing data through a weighing sensor, or the object recognition camera detects a weighing target, it is determined that a weighing event of the bulk objects has been detected.

[0032] When a weighing event of bulk items is detected, the weighing device can collect the current weighing item image through the item recognition camera, and then call the pre-trained item detection model to perform single-unit detection on the weighing item image. The item detection model can be a target detection model, which can be a model built based on a one-stage target detection algorithm such as YOLO and SSD, or a model built based on a two-stage target detection algorithm such as Faster R-CNN.

[0033] The object detection model can be obtained by labeling the weighing object images collected historically, and training the weighing object images and label data as a data set. The object detection model can also be obtained by transfer learning, by transferring the target detection model trained based on a large-scale data set to a self-built data set. Through the object detection model, the weighing equipment can detect the monomers in the weighing object image, and then obtain the object monomer detection result. Among them, the object monomer detection result may include the number of object monomers, the location data of each object monomer, and the size description data of each object monomer.

[0034] In this solution, optionally, the object detection model includes detection frame setting parameters; the detection frame setting parameters are determined by clustering algorithm analysis based on the marking frames of the object monomers in the pre-acquired training sample images.

[0035] It is easy to understand that the detection frame can be used to mark the detection target in the large image to distinguish the detection target from the background image. Depending on the detection target, the detection frame can have different shapes, which can be regular shapes such as rectangles, triangles or circles, or irregular shapes formed along the detection target. The detection frame setting parameters can be used to set parameters such as the detection frame shape, color and size. Before training the object detection model, the training sample image needs to be labeled to obtain label data. Among them, the label data can include the marking frame of each object monomer in the training sample image, and the marking frame can include information such as position, shape and size. The model trainer can perform statistical analysis on the marking frame of the object monomer in the training sample image to obtain statistical data on the shape, size and other information of the marking frame. For example, assuming that the marking frame is set to a rectangle, cluster analysis can be performed based on the size of the marking frame to obtain the size classification of the marking frame of each object monomer in the training sample image, and then the detection frame parameters can be set according to the classification results to eliminate the interference of the detection frame that is too small or too large during the object detection process and eliminate redundant detection frames.

[0036] This solution can set parameters through detection frames, which is helpful to avoid detection frame overlap and reduce detection frame redundancy from the source.

[0037] S120: Determine the density of the weighed objects according to the detection results of the individual objects.

[0038] The density of weighed items can be used to indicate the probability that the item recognition camera can completely collect all types of monomers in bulk items during weighing. If the density of weighed items is high, there may be situations where the item monomers are covered or blocked, resulting in a low monomer detection rate. The weighing device can calculate the density of weighed items based on the current item monomer detection results. For example, the weighing device can add up the areas of all item monomer detection frames, and use the ratio of the area of ​​the item monomer detection frame to the area of ​​the weighed item image as the density of weighed items.

[0039] S130: If the density of the weighed objects is lower than a preset density threshold, a single image of the object in the weighed object image is obtained, and a recognition result of the single image of the object is determined based on a pre-trained classification model.

[0040] The weighing equipment can pre-set a density threshold to evaluate whether the density of the weighed items can meet the individual detection rate requirements. If the density of the weighed items is lower than the density threshold, it means that the detected individual items can cover the categories of all bulk items and meet the detection requirements. The weighing equipment can locate the individual image of the item from the weighed item image according to the detection frame, and then input the individual image of the item into the pre-trained classification model to classify and identify each individual image of the item to obtain the recognition result of each individual image of the item. Among them, the classification model can be a classification model built based on a deep learning algorithm, such as a classification model based on a convolutional neural network. It should be noted that the individual images of the items can be of different sizes and shapes, so the classification model can read the image format when inputting the individual image of the item, and perform adaptive feature extraction and classification recognition according to the image size, shape and other attributes.

[0041] In a feasible solution, optionally, after determining the recognition result of the object single image based on the pre-trained classification model, the method further includes:

[0042] Determining the number of categories of objects in the weighing object image according to the recognition result;

[0043] If the number of categories is greater than 1, determine whether the unit prices of items in each category are the same;

[0044] If they are the same, the bulk items are weighed and settled; if they are different, classified packaging prompt information is generated according to the categories of the items in the recognition results, so as to weigh and settle the packaged items with different unit prices separately.

[0045] It is understandable that the recognition result may include information such as the category of each item. The weighing device can determine how many categories of items there are in the weighing item image, that is, the number of item categories, based on the category of each item. If the number of item categories in the weighing item image is equal to 1, it means that there is only one type of item in the weighing item image, and the weighing device can directly perform weighing and settlement. If the number of item categories in the weighing item image is greater than 1, it means that there are multiple categories of items in the weighing item image. The weighing device can further determine whether the unit prices of each category of items are consistent by retrieving the information of each category of items. If they are the same, the bulk items can be combined for weighing and settlement. If the unit prices of each category of items are inconsistent, the weighing device can generate classification and packaging prompt information based on the category of the item, for example: "There are multiple categories of items with different unit prices in the current weighing items, specifically including category A and category B. Please classify, pack and weigh. Thank you!".

[0046] This solution can determine the weighing method of bulk items based on the recognition results of the individual images of the items, which is helpful to avoid losses for customers and merchants and bring a good shopping experience to customers.

[0047] When the weighing event of bulk items is detected, the technical solution obtains the weighing item image, and performs a single-unit detection on the weighing item image based on the pre-trained item detection model; then, the density of the weighing items is determined based on the single-unit detection result of the item; when the density of the weighing items is lower than the preset density threshold, the single-unit image of the item in the weighing item image is obtained, and the recognition result of the single-unit image of the item is determined based on the pre-trained classification model. This solution can solve the problem of economic losses caused by weighing bulk items of different unit prices at the same unit price, and can accurately detect each bulk item during the weighing process, which is conducive to rapid weighing and avoiding economic losses for customers and merchants.

[0048] Embodiment 2

[0049] Figure 2A A flowchart of a bulk object identification method is provided for the second embodiment of the present invention. This embodiment is based on the above embodiment and is refined. Figure 2A As shown, the method includes:

[0050] S210: If a weighing event of bulk items is detected, an image of the weighed item collected by the weighing device is obtained, and a single-item detection is performed on the image of the weighed item based on a pre-trained item detection model to obtain a single-item detection result.

[0051] In this solution, the object detection result includes at least one detection frame information, wherein the detection frame information includes position data and range data. The position data may be data used to mark the position of the detection frame, such as the coordinates of the center point of the detection frame. The range data may be data used to describe the coverage area of ​​the detection frame, for example, when the detection frame is a rectangle, the range data may be the length and width of the rectangle.

[0052] S220: Determine the unconnected graphic area of ​​the weighing object image according to the number of detection frame information in the object single body detection result and the position data and range data in each detection frame information.

[0053] Figure 2B Schematic diagram of a method for determining the density of weighed objects provided according to the second embodiment of the present invention. Figure 2B As shown in the figure, after the single body detection of the weighing object image is performed, three detection frames are located, that is, three object single bodies are detected. Among them, each detection frame corresponds to a detection frame information, and the weighing device can define an unconnected figure to include all the detection frames, such as Figure 2B According to the number of detection frames, the position data and range data of each detection frame, the weighing device can determine the length of each side of the unconnected figure, and then calculate the area of ​​the unconnected figure.

[0054] S230, determining the density of weighing objects according to the unconnected graphic area, the number of detection frame information and the pre-calculated average area of ​​the detection frames; wherein the average area of ​​the detection frames is calculated based on the range data of each detection frame information.

[0055] The weighing equipment can calculate the density of the weighed items based on the designated unconnected graphics. For example, the area of ​​the unconnected graphics is used as the denominator, and the sum of the areas of each detection frame is used as the numerator, and the obtained area ratio is used as the density of the weighed items. Since the detection frames may have differences in shape and size during the actual detection process, in order to avoid the influence of the detection frame being too large or too small on the density calculation. The weighing equipment can calculate the average area of ​​the detection frame, and use the number of detection frame information and the average area of ​​the detection frame to calculate the effective area covered by all the detection frames. The ratio of the unconnected graphics area to the effective area is used as the density of the weighed items.

[0056] In a feasible solution, the calculation formula for the density of weighed items is:

[0057]

[0058] Among them, K represents the density of weighing objects, S represents the area of ​​unconnected graphics, N represents the number of detection frame information, and M represents the average area of ​​the detection frame.

[0059] This solution can quickly and accurately calculate the density of weighed items and avoid the influence of abnormal values ​​of the detection frame area on the density calculation.

[0060] S240: Determine whether the density of the weighed objects is higher than or equal to a preset density threshold.

[0061] If the density of the weighed objects is lower than the density threshold, S250 is executed; if the density of the weighed objects is higher than or equal to the density threshold, S260 is executed.

[0062] S250: Locate the position of each object single image according to the position data in each detection frame information.

[0063] If the density of the weighed objects is lower than the density threshold, the weighing device can locate the position of the individual image of the object in the weighed object image based on the position data, such as the midpoint coordinates of the detection frame, and then obtain the individual image of each object for classification and identification.

[0064] S260: Generate bulk object movement information according to each detection frame information, so as to prompt the user to move the bulk objects through the weighing device.

[0065] If the density of the weighed items is higher than or equal to the density threshold, the weighing device can determine the dense area based on the information of each detection frame, such as the position data of each detection frame. The weighing device can generate bulk item movement information based on the located dense area. For example, if there is a large amount of detection frame position data concentrated in area A, then area A can be considered as a dense area, and the bulk item movement information can be "items are piled up and covered in area A, please move away." The user can move the bulk items according to the prompt information to ensure that the single item detection is successfully completed. It should be noted that after detecting that the user moves the bulk items, the item recognition camera can trigger the weighing event of the bulk items again, return to execute S210, re-collect the weighing item image, and perform single item detection.

[0066] S270: Capture each object single image according to each range data and each object single image position.

[0067] It is easy to understand that the weighing device can determine each individual image of the object according to the position of each individual image of the object and the range data in the corresponding detection frame information. The weighing device can intercept each individual image of the object to perform detailed identification and classification of the individual object for the individual image of the object.

[0068] S280: Determine the recognition result of the object single image based on the pre-trained classification model.

[0069] When the weighing event of bulk items is detected, the technical solution obtains the weighing item image, and performs a single-unit detection on the weighing item image based on the pre-trained item detection model; then, the density of the weighing items is determined based on the single-unit detection result of the item; when the density of the weighing items is lower than the preset density threshold, the single-unit image of the item in the weighing item image is obtained, and the recognition result of the single-unit image of the item is determined based on the pre-trained classification model. This solution can solve the problem of economic losses caused by weighing bulk items of different unit prices at the same unit price, and can accurately detect each bulk item during the weighing process, which is conducive to rapid weighing and avoiding economic losses for customers and merchants.

[0070] Embodiment 3

[0071] Figure 3 This is a schematic diagram of the structure of a bulk article identification device provided in Embodiment 3 of the present invention. The device is configured on a weighing device; Figure 3 As shown, the device comprises:

[0072] The unit detection result generating module 310 is used to obtain the weighing object image collected by the weighing device if a weighing event of bulk items is detected, and perform unit detection on the weighing object image based on the pre-trained object detection model to obtain the unit detection result of the item;

[0073] A density determination module 320, configured to determine the density of weighed objects according to the object unit detection result;

[0074] The recognition result determination module 330 is used to obtain a single image of the weighed object in the weighed object image if the density of the weighed object is lower than a preset density threshold, and determine the recognition result of the single image of the object based on a pre-trained classification model.

[0075] In this solution, optionally, the object detection result includes at least one piece of detection frame information, wherein the detection frame information includes position data and range data;

[0076] The density determination module 320 includes:

[0077] The unconnected pattern area determination unit is used to determine the unconnected pattern area of ​​the weighing object image according to the number of detection frame information in the object single body detection result, the position data and the range data in each detection frame information;

[0078] The density determination unit is used to determine the density of the weighing objects according to the unconnected graphic area, the number of detection frame information and the pre-calculated average area of ​​the detection frame; wherein the average area of ​​the detection frame is calculated based on the range data of each detection frame information.

[0079] Based on the above solution, optionally, the calculation formula for the density of weighed items is:

[0080]

[0081] Among them, K represents the density of weighing objects, S represents the area of ​​unconnected graphics, N represents the number of detection frame information, and M represents the average area of ​​the detection frame.

[0082] In a feasible solution, optionally, the recognition result determination module 330 includes:

[0083] A single image position determination unit, configured to locate the position of each single image of each object according to the position data in each detection frame information if the density of the weighed objects is lower than a preset density threshold;

[0084] The single image capture unit is used to capture the single image of each object according to each range data and the position of the single image of each object.

[0085] In another feasible solution, optionally, the device further includes:

[0086] The movement information generating module is used to generate bulk object movement information according to each detection frame information if the density of the weighed objects is higher than or equal to a preset density threshold, so as to prompt the user to move the bulk objects through the weighing device.

[0087] In this embodiment, optionally, the device further includes:

[0088] A category quantity determination module, used to determine the category quantity of objects in the weighing object image according to the recognition result;

[0089] A unit price determination module, used to determine whether the unit prices of items in each category are the same if the number of categories is greater than 1;

[0090] The weighing and settlement module is used to weigh and settle the bulk items if they are the same; if they are different, generate classification and packaging prompt information according to the categories of the items in the recognition results, so as to weigh and settle the packaged items with different unit prices separately.

[0091] In a preferred solution, the object detection model includes detection frame setting parameters; the detection frame setting parameters are determined by clustering algorithm analysis based on the marking frames of object monomers in pre-acquired training sample images.

[0092] The bulk article identification device provided in the embodiment of the present invention can execute the bulk article identification method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0093] Embodiment 4

[0094] Figure 4 A schematic diagram of a weighing device 410 that can be used to implement an embodiment of the present invention is shown. The weighing device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The weighing device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0095] like Figure 4As shown, the weighing device 410 includes at least one processor 411, and the object identification camera 416 is connected to a memory such as a read-only memory (ROM) 412, a random access memory (RAM) 413, etc., which is communicated with the at least one processor 411, wherein the memory stores a computer program that can be executed by at least one processor, and the processor 411 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 412 or the computer program loaded from the storage unit 419 to the random access memory (RAM) 413. In the RAM 413, various programs and data required for the operation of the weighing device 410 can also be stored. The processor 411, the ROM 412, and the RAM 413 are connected to each other via a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.

[0096] A number of components in the weighing device 410 are connected to the I / O interface 415, including: an object recognition camera 416 for collecting images of the weighing object; an input unit 417, such as a keyboard, a mouse, etc.; an output unit 418, such as various types of displays, speakers, etc.; a storage unit 419, such as a disk, an optical disk, etc.; and a communication unit 420, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 420 allows the weighing device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0097] The processor 411 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 411 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 411 performs the various methods and processes described above, such as a bulk item recognition method.

[0098] In some embodiments, the bulk item identification method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 419. In some embodiments, part or all of the computer program may be loaded and / or installed on the weighing device 410 via the ROM 412 and / or the communication unit 420. When the computer program is loaded into the RAM 413 and executed by the processor 411, one or more steps of the bulk item identification method described above may be performed. Alternatively, in other embodiments, the processor 411 may be configured to perform the bulk item identification method in any other appropriate manner (e.g., by means of firmware).

[0099] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0100] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0101] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0102] To provide interaction with a user, the systems and techniques described herein may be implemented on a bulk item identification system having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) through which a user can provide input to the bulk item identification system. Other types of devices may also be used to provide interaction with a user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0103] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0104] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0105] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0106] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying bulk items, characterized in that: The method is performed by a weighing device, and the method comprises: If a weighing event of bulk items is detected, an image of the weighed item collected by the weighing device is obtained, and a single-item detection is performed on the image of the weighed item based on a pre-trained item detection model to obtain a single-item detection result; Determining the density of weighed objects according to the detection results of the individual objects; If the density of the weighed objects is lower than the preset density threshold, the individual image of the object in the weighed object image is obtained, and the recognition result of the individual image of the object is determined based on the pre-trained classification model; wherein the preset density threshold is used to evaluate whether the density of the weighed objects meets the individual detection rate requirement.

2. The method according to claim 1, characterized in that: The object detection result includes at least one piece of detection frame information, wherein the detection frame information includes position data and range data; Determining the density of weighed objects according to the object monomer detection results includes: Determine the circumscribed rectangular area of ​​the weighing object image according to the number of detection frame information in the object single body detection result and the position data and range data in each detection frame information; The density of the weighed objects is determined according to the area of ​​the circumscribed rectangle, the number of the detection frame information and the pre-calculated average area of ​​the detection frame; wherein the average area of ​​the detection frame is calculated based on the range data of each detection frame information.

3. The method according to claim 2, characterized in that The calculation formula for the density of weighed items is: ; in, Indicates the density of weighed items. represents the area of ​​the circumscribed rectangle, Indicates the number of detection box information, Represents the average area of ​​the detection box.

4. The method according to claim 2, characterized in that: If the density of the weighed objects is lower than a preset density threshold, obtaining a single image of the object in the weighed object image includes: If the density of the weighed objects is lower than a preset density threshold, the position of each object single image is located according to the position data in each detection frame information; According to each range data and each object single image position, each object single image is captured.

5. The method according to claim 2, characterized in that: After determining the density of the weighed objects, the method further includes: If the density of the weighed objects is higher than or equal to a preset density threshold, bulk object movement information is generated according to the information of each detection frame, so as to prompt the user to move the bulk objects through the weighing device.

6. The method according to claim 1, characterized in that After determining the recognition result of the object single image based on the pre-trained classification model, the method further includes: Determining the number of categories of objects in the weighing object image according to the recognition result; If the number of categories is greater than 1, determine whether the unit prices of items in each category are the same; If they are the same, the bulk items are weighed and settled; if they are different, classified packaging prompt information is generated according to the categories of the items in the recognition results, so as to weigh and settle the packaged items with different unit prices separately.

7. The method according to claim 1, characterized in that The object detection model includes detection frame setting parameters; the detection frame setting parameters are determined by clustering algorithm analysis based on the marking frames of object monomers in the pre-acquired training sample images.

8. A bulk object identification device, characterized in that: The device is configured on a weighing device, and comprises: A single-unit detection result generating module is used to obtain a weighing object image collected by the weighing device if a weighing event of a bulk object is detected, and to perform single-unit detection on the weighing object image based on a pre-trained object detection model to obtain a single-unit object detection result; A density determination module, used to determine the density of weighed objects according to the detection results of the object monomers; The recognition result determination module is used to obtain the individual image of the object in the image of the weighed object if the density of the weighed object is lower than the preset density threshold, and determine the recognition result of the individual image of the object based on a pre-trained classification model; wherein the preset density threshold is used to evaluate whether the density of the weighed object meets the individual detection rate requirement.

9. A weighing device, characterized in that: The device includes: an object recognition camera, at least one processor and a memory; The object recognition camera is communicatively connected to the processor, and the processor is communicatively connected to the memory; Wherein, the object recognition camera is used to collect images of weighing objects; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so as to enable the at least one processor to perform the bulk article identification method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the bulk article identification method according to any one of claims 1 to 7 when executed.