Logistics package management method and system, readable storage medium and computer
By real-time acquisition and frequency domain processing of the image data of logistics packages, combined with the convolutional neural network to identify the package type, the problem of not being able to identify abnormal targets in the existing technology is solved, and efficient classification management and identification of logistics packages is achieved.
Patent Information
- Application Number
- CN202411112832.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing logistics parcel management methods cannot effectively identify and distinguish abnormal goals in logistics transportation, resulting in the inability to achieve rapid management and classification.
By collecting image data of logistics packages in real time, performing image preprocessing and frequency domain conversion, building target areas for instance segmentation, combining convolutional neural network for binary classification, identifying package types and classifying management.
It improves the accuracy and efficiency of item identification, realizes effective classification management of logistics packages, and enhances abnormal target detection capabilities.
Smart Images

Figure CN118968484A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of e-commerce, and in particular to a logistics package management method, system, readable storage medium and computer. Background Art
[0002] With the rapid development of science and technology and the improvement of people's living standards, more and more people are starting to shop through e-commerce platforms. As the number of shopping increases, the amount of express logistics is also increasing. New management methods represented by multi-dimensional, all-round real-time monitoring have gradually penetrated into every level of logistics management, and have had a profound impact on the logistics management model.
[0003] In the process of logistics transportation, as the number of logistics items increases, during the handling of logistics items, the phenomenon of items falling when the packages are transported on the conveyor belt also often occurs. However, the current logistics package management method is mainly used to identify the items in logistics transportation, and it is impossible to further distinguish the items, and it is also impossible to detect abnormal targets in the items. For example, the traditional target detection method cannot directly detect abnormal targets, but can only identify all express cartons in the conveyor belt area, and cannot achieve effective distinction and rapid management. Summary of the invention
[0004] Based on this, the purpose of the present invention is to provide a logistics package management method, system, readable storage medium and computer to at least solve the deficiencies in the above-mentioned technology.
[0005] The present invention provides a logistics package management method, comprising: Acquire an image data set of the logistics package in the active window in real time, and perform image preprocessing on the image data set to convert the image data set into a frequency domain image set; Constructing a target region, and performing instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; Calculating the degree of overlap between each of the instance segmentation results and the target area, and determining corresponding input data based on the degree of overlap and a preset degree of overlap threshold; Inputting the input data into the convolutional neural network that has completed learning to perform binary classification solution to obtain solution data of the input data; The package type of the logistics package corresponding to the input data is determined according to the solution data, and the logistics package is classified and managed according to the package type.
[0006] Furthermore, the step of performing image preprocessing on the image data set to convert the image data set into a frequency domain image set includes: Performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; The binary image data set is subjected to Fourier transformation so as to convert the binary image data set into a frequency domain image set in a complex frequency domain.
[0007] Furthermore, the steps of constructing the target area include: Acquire standard image data containing a plurality of gray levels, and calculate a gray level histogram corresponding to the standard image data; Performing background differentiation on the grayscale histogram and filtering out background pixels in the grayscale histogram to obtain target image data; The device parameters of the acquisition device are obtained, and the position data of the acquisition device and the target package in the target image data are calculated according to the device parameters, and the target area is constructed based on the position data.
[0008] Furthermore, the step of inputting the input data into the convolutional neural network that has completed learning to perform binary classification solution to obtain solution data of the input data includes: Dividing the input data into abnormal data and normal data, and inputting the abnormal data and the normal data into the convolutional neural network that has completed learning respectively; Target detection is performed on the abnormal data and the regular data based on the fully connected layer in the convolutional neural network to obtain abnormal output data and regular output data.
[0009] Further, the step of determining the package type of the logistics package corresponding to the input data according to the solution data includes: Using a deep learning algorithm to perform probability calculation on the abnormal output data and the normal output data to obtain corresponding abnormal output probabilities and normal output probabilities; The package type of the logistics package corresponding to the input data is determined according to the abnormal output probability and the regular output probability, wherein the package type includes an abnormal target type and a regular target type.
[0010] Furthermore, the expression of the package type is:
[0011] In the formula, Indicates the exception target type, Indicates the general target type, represents the probability of abnormal output, represents the general output probability, represents the weight parameters of the network, Indicates Input data.
[0012] The present invention also provides a logistics package management system, comprising: A data acquisition module, used for acquiring an image data set of the logistics package in the active window in real time, and performing image preprocessing on the image data set to convert the image data set into a frequency domain image set; An instance segmentation module is used to construct a target region and perform instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; A coincidence calculation module, used to calculate the coincidence between each of the instance segmentation results and the target area, and determine the corresponding input data based on the coincidence and a preset coincidence threshold; A data solving module, used for inputting the input data into the convolutional neural network that has completed learning to perform binary classification solving, so as to obtain solving data of the input data; The package classification module is used to determine the package type of the logistics package corresponding to the input data according to the solution data, and classify and manage the logistics package according to the package type.
[0013] Furthermore, the data acquisition module includes: An image processing unit, used for performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; The data conversion unit is used to perform Fourier transform on the binary image data set so as to convert the binary image data set into a frequency domain image set in a complex frequency domain.
[0014] Furthermore, the instance segmentation module includes: A grayscale processing unit, used to obtain standard image data containing a plurality of grayscale levels, and calculate a grayscale histogram corresponding to the standard image data; A background screening unit, used for performing background differentiation on the grayscale histogram and screening out background pixels in the grayscale histogram to obtain target image data; The area construction unit is used to obtain device parameters of the acquisition device, calculate the position data of the acquisition device and the target package in the target image data according to the device parameters, and construct the target area based on the position data.
[0015] Furthermore, the data solving module includes: A data division unit, used for dividing the input data into abnormal data and normal data, and inputting the abnormal data and the normal data into the convolutional neural network that has completed learning respectively; A target detection unit is used to perform target detection on the abnormal data and the regular data based on the fully connected layer in the convolutional neural network to obtain abnormal output data and regular output data.
[0016] Furthermore, the package classification module includes: A probability calculation unit, used to perform probability calculation on the abnormal output data and the normal output data using a deep learning algorithm to obtain corresponding abnormal output probabilities and normal output probabilities; A package classification unit is used to determine the package type of the logistics package corresponding to the input data according to the abnormal output probability and the normal output probability, wherein the package type includes an abnormal target type and a normal target type.
[0017] The present invention also provides a readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned logistics package management method is implemented.
[0018] The present invention also proposes a computer, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned logistics package management method when executing the computer program.
[0019] The logistics package management method, system, readable storage medium and computer of the present invention can make data acquisition more immediate by real-time acquisition of the image data set of the logistics package in the active window, and use the target area to perform instance segmentation processing on the frequency domain image set converted from the image data set, so as to make the identification of regional items more convenient and improve the accuracy and efficiency of item identification; compare the instance segmentation result with the overlap of the target area to determine the corresponding input data, improve the accuracy of item identification through further detection, and use the input data to identify the package type of the logistics package, thereby realizing the classification management of the logistics package. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 is a flow chart of a logistics package management method in a first embodiment of the present invention; Figure 2 for Figure 1 Detailed flow chart of step S101; Figure 3 for Figure 1 Detailed flow chart of step S102; Figure 4 for Figure 1 Detailed flow chart of step S104; Figure 5 for Figure 1 Detailed flow chart of step S105; Figure 6 is a structural block diagram of a logistics package management system in a second embodiment of the present invention; Figure 7FIG. 4 is a structural block diagram of a computer in a third embodiment of the present invention.
[0021] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION
[0022] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.
[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.
[0024] Embodiment 1 See also Figure 1 , which shows a logistics package management method in a first embodiment of the present invention, and the method specifically includes steps S101 to S105: S101, collecting an image data set of a logistics package in an active window in real time, and performing image preprocessing on the image data set to convert the image data set into a frequency domain image set; For further information, see Figure 2 , the step S101 specifically includes steps S1011~S1012: S1011, performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; S1012: Perform Fourier transform on the binary image data set to convert the binary image data set into a frequency domain image set in a complex frequency domain.
[0025] In the specific implementation, the parcels will be sorted before logistics distribution, so that the parcels are sent to various places from the transit warehouse. The parcels are collected in real time by two image acquisition devices (one on the left and one on the right) installed at the corresponding sorting window (i.e., the active window). After the corresponding image data set is collected, in order to better extract the contour of the collected parcel, it is necessary to perform image filtering and grayscale changes on the image data set to obtain the binary image data set corresponding to the image data set. The pixel points in the image data set are scanned by matrix operation (wherein the selection of pixel points is determined by the convolution kernel size of the matrix operation, for example: when the convolution kernel size is 6*6, the pixel points are 36 pixels), the pixels are sorted and the median is taken, and the middle value corresponding to the kernel is replaced with the median, thereby realizing filtering processing; Specifically, the grayscale transformation is performed on the filtered image to enhance the information of the wrapped area in the image to reduce the generation of interference information; the image data after the grayscale transformation (binary image data) is Fourier transformed to convert the binary image data into a frequency domain image set in the complex frequency domain.
[0026] S102, constructing a target region, and performing instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; For further information, see Figure 3 , the step S102 specifically includes steps S1021 to S1023: S1021, acquiring standard image data containing a plurality of gray levels, and calculating a gray level histogram corresponding to the standard image data; S1022, performing background differentiation on the grayscale histogram, and filtering out background pixels in the grayscale histogram to obtain target image data; S1023, acquiring device parameters of a collection device, and calculating position data of the collection device and a target package in the target image data according to the device parameters, and constructing a target area based on the position data.
[0027] In a specific implementation, standard image data containing several gray levels is obtained, wherein the standard image data is image data that has completed classification management, a gray histogram corresponding to the standard image data is calculated, and after normalization processing, the histogram corresponds to the probability of each gray level in the image, the histogram is divided into background and foreground for the gray threshold, and the variance of the background and foreground is calculated; it is determined whether all pixels in the image are background, and the background pixels are screened out from the image, thereby obtaining the target image data.
[0028] Furthermore, the device parameters of the above-mentioned image acquisition device are obtained, wherein the device parameters include the device position, the distance between the two devices and the focal length. The edge of the target package in the standard image data is detected using the Otsu binary segmentation method and the device position, the distance between the two devices and the focal length, thereby obtaining the position data of the target package, and constructing the target area based on the position data. In this embodiment, different packages have different sizes and corresponding delivery areas, and their corresponding target areas are also different. The larger the package size, the larger the target area, and the different delivery areas have different target areas.
[0029] Specifically, after obtaining the target area, instance segmentation processing is performed on the frequency domain image set obtained above based on the target area. Instance segmentation is to effectively identify the boundaries of target detection to mark different individuals of the same object in the collected image data. Since there may be multiple packages in the image data collected by the image acquisition device, instance segmentation can effectively identify the boundaries of the packages.
[0030] S103, calculating the degree of overlap between each of the instance segmentation results and the target area, and determining corresponding input data based on the degree of overlap and a preset degree of overlap threshold; In a specific implementation, when a package is smaller in size, the boundary obtained after instance segmentation processing should be smaller. Therefore, by inputting the instance segmentation processing result and the target area obtained above into the preset deep learning model for overlap calculation, it is possible to effectively identify whether the package belongs to the package type corresponding to the current target area, and mark the overlapping area with an overlap lower than a preset overlap threshold (in this embodiment, the overlap threshold is 80%) as input data. When the overlap is lower than the preset overlap threshold, it means that the package may be abnormal, so it needs to be analyzed and processed to ensure the package transportation.
[0031] S104, inputting the input data into the convolutional neural network that has completed learning to perform binary classification solution to obtain solution data of the input data; For further information, see Figure 4 , the step S104 specifically includes steps S1041-S1042: S1041, dividing the input data into abnormal data and normal data, and inputting the abnormal data and the normal data into the convolutional neural network that has completed learning respectively; S1042, performing target detection on the abnormal data and the regular data based on the fully connected layer in the convolutional neural network to obtain abnormal output data and regular output data.
[0032] In the specific implementation, the input data obtained above is divided into abnormal data and regular data, and the abnormal data and regular data are input into the convolutional neural network that has completed learning, and the fully connected layer of the convolutional neural network is used to perform target detection, and the target detection is marked as a two-class prediction problem, that is, the prediction result calculated by the final convolutional neural network model contains two types of targets: abnormal output data and regular output data.
[0033] S105, determining the package type of the logistics package corresponding to the input data according to the solution data, and classifying and managing the logistics package according to the package type.
[0034] For further information, see Figure 5 , the step S105 specifically includes steps S1051~S1052: S1051, using a deep learning algorithm to perform probability calculation on the abnormal output data and the normal output data to obtain corresponding abnormal output probabilities and normal output probabilities; S1052: Determine the package type of the logistics package corresponding to the input data according to the abnormal output probability and the normal output probability, wherein the package type includes an abnormal target type and a normal target type.
[0035] In the specific implementation, the probability of abnormal output data and regular output data is calculated through the deep learning algorithm, that is, the probability of the current package data being abnormal output data or regular output data is calculated, and the activation value of the abnormal output data is calculated through the activation function in the deep learning algorithm. And the activation value of the conventional output data ,in, represents the weight parameters of the network, Represents the data input to the activation function.
[0036] Further, using softmax The activation value of the function for the abnormal output data obtained above And the activation value of the conventional output data Normalize and convert it into the probability of belonging to different target categories, and according to softmax The function calculates the two probabilities to obtain the corresponding probability value, thereby determining the package type of the logistics package corresponding to the input data. After obtaining the corresponding package type, the logistics package will be classified and managed according to the package type, where the expression of the package type is:
[0037] In the formula, Indicates the exception target type, Indicates the general target type, represents the probability of abnormal output, represents the general output probability, represents the weight parameters of the network, Indicates Input data.
[0038] In summary, the logistics package management method in the above-mentioned embodiment of the present invention can make data acquisition more immediate by real-time acquisition of the image data set of the logistics package in the active window, and use the target area to perform instance segmentation processing on the frequency domain image set converted from the image data set, thereby making the identification of regional objects more convenient and improving the accuracy and efficiency of object identification; comparing the instance segmentation result with the overlap of the target area to determine the corresponding input data, improving the accuracy of object identification through further detection, and using the input data to identify the package type of the logistics package, thereby realizing the classification management of the logistics package.
[0039] Embodiment 2 Another aspect of the present invention is to provide a logistics package management system, please refer to Figure 6 , which shows a logistics package management system in a second embodiment of the present invention, the system includes: The data acquisition module 11 is used to acquire an image data set of the logistics package in the active window in real time, and perform image preprocessing on the image data set to convert the image data set into a frequency domain image set; Furthermore, the data acquisition module 11 includes: An image processing unit, used for performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; The data conversion unit is used to perform Fourier transform on the binary image data set so as to convert the binary image data set into a frequency domain image set in a complex frequency domain.
[0040] An instance segmentation module 12 is used to construct a target region and perform instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; Furthermore, the instance segmentation module 12 includes: A grayscale processing unit, used to obtain standard image data containing a plurality of grayscale levels, and calculate a grayscale histogram corresponding to the standard image data; A background screening unit, used for performing background differentiation on the grayscale histogram and screening out background pixels in the grayscale histogram to obtain target image data; The area construction unit is used to obtain device parameters of the acquisition device, calculate the position data of the acquisition device and the target package in the target image data according to the device parameters, and construct the target area based on the position data.
[0041] The overlap calculation module 13 is used to calculate the overlap between each of the instance segmentation results and the target area, and determine the corresponding input data based on the overlap and a preset overlap threshold; A data solving module 14 is used to input the input data into the convolutional neural network that has completed learning to perform binary classification solving, so as to obtain solving data of the input data; Furthermore, the data solving module 14 includes: A data division unit, used for dividing the input data into abnormal data and normal data, and inputting the abnormal data and the normal data into the convolutional neural network that has completed learning respectively; A target detection unit is used to perform target detection on the abnormal data and the regular data based on the fully connected layer in the convolutional neural network to obtain abnormal output data and regular output data.
[0042] The package classification module 15 is used to determine the package type of the logistics package corresponding to the input data according to the solution data, and classify and manage the logistics package according to the package type.
[0043] Furthermore, the package classification module 15 includes: A probability calculation unit, used to perform probability calculation on the abnormal output data and the normal output data using a deep learning algorithm to obtain corresponding abnormal output probabilities and normal output probabilities; A package classification unit is used to determine the package type of the logistics package corresponding to the input data according to the abnormal output probability and the normal output probability, wherein the package type includes an abnormal target type and a normal target type.
[0044] The functions or operation steps implemented when the above modules and units are executed are generally the same as those in the above method embodiments, and will not be repeated here.
[0045] The implementation principle and technical effects of the logistics parcel management system provided in the embodiment of the present invention are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the system embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.
[0046] Embodiment 3 The present invention also provides a computer, see Figure 7, shown is a computer in the third embodiment of the present invention, including a memory 10, a processor 20, and a computer program 30 stored in the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, the above-mentioned logistics package management method is implemented.
[0047] The memory 10 includes at least one type of readable storage medium, which includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 10 may be an internal storage unit of a computer, such as a hard disk of the computer. In other embodiments, the memory 10 may also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Further, the memory 10 may also include both an internal storage unit of the computer and an external storage device. The memory 10 may be used not only to store application software and various types of data installed in the computer, but also to temporarily store data that has been output or is to be output.
[0048] Among them, in some embodiments, the processor 20 can be an electronic control unit (Electronic Control Unit, abbreviated as ECU, also known as a vehicle computer), a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor or other data processing chip, used to run the program code stored in the memory 10 or process data, such as executing access restriction programs, etc.
[0049] It should be pointed out that Figure 7 The structure shown does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.
[0050] The embodiment of the present invention further provides a readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the logistics package management method as described above is implemented.
[0051] Those skilled in the art will appreciate that the logic and / or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, "computer-readable medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0052] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0053] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0054] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0055] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.
Claims
1. A logistics package management method, characterized in that: include: Acquire an image data set of the logistics package in the active window in real time, and perform image preprocessing on the image data set to convert the image data set into a frequency domain image set; Constructing a target region, and performing instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; Calculating the degree of overlap between each of the instance segmentation results and the target area, and determining corresponding input data based on the degree of overlap and a preset degree of overlap threshold; Inputting the input data into the convolutional neural network that has completed learning to perform binary classification solution to obtain solution data of the input data; The package type of the logistics package corresponding to the input data is determined according to the solution data, and the logistics package is classified and managed according to the package type.
2. The logistics package management method according to claim 1, characterized in that: The step of performing image preprocessing on the image data set to convert the image data set into a frequency domain image set comprises: Performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; The binary image data set is subjected to Fourier transformation so as to convert the binary image data set into a frequency domain image set in a complex frequency domain.
3. The logistics package management method according to claim 1, characterized in that: The steps to construct the target area include: Acquire standard image data containing a plurality of gray levels, and calculate a gray level histogram corresponding to the standard image data; Performing background differentiation on the grayscale histogram and filtering out background pixels in the grayscale histogram to obtain target image data; The device parameters of the acquisition device are obtained, and the position data of the acquisition device and the target package in the target image data are calculated according to the device parameters, and the target area is constructed based on the position data.
4. The logistics package management method according to claim 1, characterized in that: The step of inputting the input data into the convolutional neural network that has completed learning to perform binary classification solution to obtain solution data of the input data includes: Dividing the input data into abnormal data and normal data, and inputting the abnormal data and the normal data into the convolutional neural network that has completed learning respectively; Target detection is performed on the abnormal data and the regular data based on the fully connected layer in the convolutional neural network to obtain abnormal output data and regular output data.
5. The logistics package management method according to claim 4, characterized in that: The step of determining the package type of the logistics package corresponding to the input data according to the solution data comprises: Using a deep learning algorithm to perform probability calculation on the abnormal output data and the normal output data to obtain corresponding abnormal output probabilities and normal output probabilities; The package type of the logistics package corresponding to the input data is determined according to the abnormal output probability and the regular output probability, wherein the package type includes an abnormal target type and a regular target type.
6. The logistics package management method according to claim 5, characterized in that: The expression of the package type is: In the formula, Indicates the exception target type, Indicates the general target type, represents the probability of abnormal output, represents the general output probability, represents the weight parameters of the network, Indicates Input data.
7. A logistics package management system, characterized in that: include: A data acquisition module, used for acquiring an image data set of the logistics package in the active window in real time, and performing image preprocessing on the image data set to convert the image data set into a frequency domain image set; An instance segmentation module is used to construct a target region and perform instance segmentation processing on the frequency domain image set based on the target region to obtain a corresponding instance segmentation result; A coincidence calculation module, used to calculate the coincidence between each of the instance segmentation results and the target area, and determine the corresponding input data based on the coincidence and a preset coincidence threshold; A data solving module, used for inputting the input data into the convolutional neural network that has completed learning to perform binary classification solving, so as to obtain solving data of the input data; The package classification module is used to determine the package type of the logistics package corresponding to the input data according to the solution data, and classify and manage the logistics package according to the package type.
8. The logistics package management system according to claim 7, characterized in that: The data acquisition module comprises: An image processing unit, used for performing image filtering and grayscale transformation on the image data set to obtain a corresponding binary image data set; The data conversion unit is used to perform Fourier transform on the binary image data set so as to convert the binary image data set into a frequency domain image set in a complex frequency domain.
9. A readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the logistics package management method as described in any one of claims 1 to 6 is implemented.
10. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the logistics package management method as described in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Instance segmentation network training model and training method, and instance segmentation network
CN112001401A
Discrete cosine transform (DCT)-based mask representation instance segmentation method
CN113223038A
Complex scene task-oriented space manipulator grabbing sensing method
CN116476059A
Space target key part detection and identification method, device and equipment based on vision
CN116935364A
