Design method of visual high-speed parcel single-piece separation modular distributed system

Abstract

The invention discloses a visual high-speed parcel single-piece separation modular distributed system design method, and relates to the technical field of computer vision, and the method comprises the steps: setting the row and column distribution and size values of a single-piece separation module according to system configuration parameters; generating an independent coordinate system corresponding to each camera according to the detection range of the camera, generating absolute position coordinates of a detection target of each camera through a visual identification algorithm, and performing unified compensation coordinate conversion to generate absolute coordinates of all detection targets; absolute coordinates of all detection targets are effectively filtered through a filter, repeated coordinates are eliminated and input into a separation strategy algorithm module of the system, and parcel single-piece separation is achieved. Modular deployment is realized by adopting a single-piece separation module, flexible expansion or reduction is supported to adapt to different logistics scene requirements, the real-time performance, the detection precision and the separation stability of the system are remarkably improved through multi-camera distributed calculation, a compensation algorithm and a filter, and the hardware requirement and the deployment cost are reduced.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to a design method for a visual high-speed package single-piece separation modular distributed system. Background Art

[0002] In the field of modern logistics, warehousing and distribution, with the rapid development of e-commerce and express delivery industries, logistics sorting systems need to handle the growing volume of packages and meet the needs of efficient and accurate sorting. Traditional single-piece separation systems often have the following problems: non-standard equipment leads to insufficient scalability. Common single-piece separation system equipment specifications are fixed. When processing packages of different specifications and quantities, the system structure needs to be redesigned or adjusted, resulting in a long deployment cycle and high cost; the computational burden of traditional stitching images. In multi-camera systems, in order to achieve package positioning in a unified coordinate system, image stitching technology is usually used. However, this method requires alignment and stitching of images taken by multiple cameras, which has high computational complexity and is prone to system delays and instability; high requirements for the deployment of multi-camera systems. Traditional methods require that the installation positions of cameras be absolutely parallel, otherwise it will affect the quality of image stitching. This strict installation requirement not only increases the difficulty of deployment, but also places high demands on hardware and debugging capabilities; repeated detection and missed detection problems. In the boundary area of ​​multiple cameras, packages may be repeatedly detected by multiple cameras, resulting in duplicate or conflicting coordinates, which in turn affects the execution effect of the separation strategy. Summary of the invention

[0003] Based on the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a visual high-speed packaging single-piece separation modular distributed system design method to solve the above-mentioned technical problems.

[0004] To achieve the above object, the present invention provides the following technical solution: a method for designing a modular distributed system for high-speed visual package separation, wherein the system is provided with a single-piece separation module of uniform size, and the method comprises:

[0005] Set the row and column distribution and size values ​​of the single separation module according to the system configuration parameters;

[0006] According to the detection range of each camera, an independent coordinate system corresponding to each camera is generated, and the same visual recognition algorithm is entered to generate the absolute position coordinates of the detection target of each camera;

[0007] By performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, the absolute coordinates of all detection targets are generated;

[0008] Effectively filtering the absolute coordinates of all the detected targets through a filter to eliminate duplicate coordinates;

[0009] The absolute coordinates of all detected targets after eliminating duplicate coordinates are input into the separation strategy algorithm module of the system to achieve single-piece separation of packages.

[0010] The present invention is further configured such that the generation of the independent coordinate system is based on the detection range of each camera and is automatically generated by inputting the row and column distribution of the single separation module and the size parameters of each module.

[0011] The present invention is further configured that the visual recognition algorithm uses each camera to independently obtain the detection image, and through the target visual processing based on the convolutional neural network, the relative position of the package in the independent coordinate system is independently calculated in a distributed manner. Under the condition of ensuring the uniformity of the timestamp, a virtual unified coordinate system is realized through the visual recognition algorithm to avoid obtaining the unified coordinate system by splicing images, wherein the real-time true position (X) of the package of each camera is calculated separately at the same time t0. i,t0 , Y j,t0 ).

[0012] The present invention is further configured to generate the absolute coordinates of all detection targets by performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, including:

[0013] Set the coordinate values ​​of the first group of cameras on the left to the absolute coordinate values ​​in the unified coordinate system;

[0014] For cameras other than the first group of cameras on the left, compensate the target coordinate values ​​in their respective coordinate systems and convert them into absolute coordinates in the unified coordinate system required by the separation algorithm. The compensated coordinates are (X i+(i-1)Δd,t0 , Y j,t0 ), where Δd is the relative displacement between cameras.

[0015] The present invention is further configured to effectively filter the absolute coordinates of all the detected targets through a filter to eliminate duplicate coordinates, including:

[0016] The width of the camera boundary area is set to be larger than the maximum size of the package in the system to ensure that the package is effectively detected by only one camera in the boundary area. The camera on the left side of the boundary area is set as camera 1, and the camera on the right side of the boundary area is set as camera 2.

[0017] When the package tail coordinate X ib >X1, the coordinates of camera 1 are marked as invalid and not uploaded to the separation strategy algorithm module, while the coordinates of camera 2 are marked as valid and start to be uploaded;

[0018] When the package tail coordinate X ib ≤X1, the coordinates of the second camera are marked as invalid and not uploaded to the separation strategy algorithm module. ibis the horizontal coordinate of the tail of the camera No. 1 package, X1 is the horizontal coordinate of the boundary point of the intersection area between camera No. 1 and camera No. 2. If the detection area of ​​camera No. 1 is from X start To X end , and the width of the boundary area between camera 1 and camera 2 is L, X1=X end -L.

[0019] The present invention is further configured such that the size of the single-piece separation module is of standardized design, the length and width of the module are set to 600 mm and 150 mm, and the splicing or expansion of multiple modules is supported according to different needs, and the camera is arranged directly above the single-piece separation module.

[0020] The present invention is further configured such that the separation strategy algorithm module calculates the separation path of each package in real time according to the absolute coordinates of all detection targets, and guides the package to the target separation outlet through the actuator of the control system.

[0021] The present invention is further configured to add a camera to the human-machine system after setting the system configuration parameters, specify the detection range, and realize modular increase or decrease of the single separation module and the camera, so that the packages can be separated quickly and efficiently.

[0022] The present invention is further configured such that the method can adapt to the detection requirements of packages of different sizes by dynamically adjusting the width and boundary coordinates of the camera boundary area.

[0023] The present invention is further configured such that the target visual processing based on the convolutional neural network includes a densely packed visual real-time precise detection algorithm based on the YOLOv5n network.

[0024] The present invention provides a modular distributed system design method for visual high-speed package single-piece separation, which sets the row and column distribution and size value of the single-piece separation module according to the system configuration parameters; generates an independent coordinate system corresponding to each camera according to the set detection range of each camera, enters the same visual recognition algorithm, and generates the absolute position coordinates of the detection target of each camera; performs unified compensation coordinate conversion on the absolute position coordinates of the detection target of each camera to generate the absolute coordinates of all detection targets; effectively filters the absolute coordinates of all detection targets through a filter to eliminate duplicate coordinates; inputs the absolute coordinates of all detection targets after eliminating duplicate coordinates into the separation strategy algorithm module of the system to achieve package single-piece separation, and the beneficial effects produced include:

[0025] 1. Improve the flexibility and adaptability of the system: By adopting a single-piece separation module of uniform size, the modular deployment of the system is realized, which supports flexible expansion or reduction of the number of separation modules, can quickly adapt to the logistics scene requirements of different specifications and processing volumes, and reduce the time and cost of redesigning and adjusting the system;

[0026] 2. Eliminate the performance bottleneck of traditional image stitching methods: Adopt a multi-camera distributed computing method to replace the traditional image stitching method. By generating an independent coordinate system for each camera and processing the relative coordinates of the target package separately based on a convolutional neural network, a unified coordinate system mapping is achieved through a compensation algorithm. This avoids the high demand for hardware and computing resources for image stitching, significantly reduces system latency, and improves real-time performance and computing efficiency.

[0027] 3. Higher detection accuracy in the camera interface area: A filter is set in the camera interface area. Through the coordinate switching mechanism and the compensation algorithm, the coordinate conflict caused by repeated package detection is effectively eliminated, ensuring that each package is uploaded with valid coordinates by only one camera. The width of the interface area is designed to be larger than the maximum size of the package, which fundamentally avoids the problem of missed detection or repeated detection and enhances the stability of the separation strategy.

[0028] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing 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. In the drawings:

[0030] Figure 1 A flowchart of a method for designing a modular distributed system for high-speed visual package separation of a single piece is shown as an exemplary embodiment of the present invention;

[0031] Figure 2 An implementation diagram of a traditional image stitching method shown as an exemplary embodiment of the present invention;

[0032] Figure 3 An independent coordinate system diagram corresponding to each camera of a modular distributed system design method for high-speed visual package separation of a single piece is shown as an exemplary embodiment of the present invention;

[0033] Figure 4 A unified compensation coordinate transformation diagram of a modular distributed system design method for high-speed visual package separation of a single piece is shown as an exemplary embodiment of the present invention;

[0034] Figure 5A filter diagram of a modular distributed system design method for visual high-speed package single-piece separation is shown as an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0035] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0036] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0037] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0038] A method for designing a modular distributed system for high-speed visual package separation of single pieces, wherein the system is provided with single piece separation modules of uniform size, and the method is as follows Figure 1 As shown, including:

[0039] Set the row and column distribution and size values ​​of the single separation module according to the system configuration parameters;

[0040] According to the detection range of each camera, an independent coordinate system corresponding to each camera is generated, and the same visual recognition algorithm is entered to generate the absolute position coordinates of the detection target of each camera;

[0041] By performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, the absolute coordinates of all detection targets are generated;

[0042] Effectively filtering the absolute coordinates of all the detected targets through a filter to eliminate duplicate coordinates;

[0043] The absolute coordinates of all detected targets after eliminating duplicate coordinates are input into the separation strategy algorithm module of the system to achieve single-piece separation of packages.

[0044] Specifically, the single-piece separation module is the core component of the entire system to achieve single-piece separation of packages. The single-piece separation module provides a standardized physical platform, which is responsible for carrying and transporting packages, so that the packages can flow along the preset path during the separation process, thereby achieving efficient single-piece separation.

[0045] The present invention is further configured such that the generation of the independent coordinate system is based on the detection range of each camera, and is automatically generated by inputting the row and column distribution of the single separation module and the size parameters of each module. Specifically, the detection range of each camera is the effective area of ​​its field of view. The range is determined by the position of the camera, the viewing angle and the size of the single separation module. The detection range of each camera is a rectangular area, and the row and column distribution of the separation modules in the system directly determines the area that each camera needs to cover. Detection range = [i×module width, (i+1)×module width]×[j×module length, (j+1)×module length], where i and j are the row and column indexes of the module; according to the detection range covered by each camera, the system generates an independent coordinate system for each camera: X-axis: corresponding to the width direction of the module. Y-axis: corresponding to the length direction of the module. The independent coordinate system of each camera strictly corresponds to its detection range to form a relative coordinate system; by automatically generating an independent coordinate system, the detection range of each camera is clearly defined, which helps to improve the accuracy of target detection and avoid duplication and omission of detection areas. Simply input basic parameters to automatically generate an independent coordinate system without the need to manually divide the detection range, significantly shortening system deployment and debugging time.

[0046] The present invention is further configured that the visual recognition algorithm uses each camera to independently obtain the detection image, and through the target visual processing based on the convolutional neural network, the relative position of the package in the independent coordinate system is independently calculated in a distributed manner. Under the condition of ensuring the uniformity of the timestamp, a virtual unified coordinate system is realized through the visual recognition algorithm to avoid obtaining the unified coordinate system by splicing images, wherein the real-time true position (X) of the package of each camera is calculated separately at the same time t0. i,t0 , Y j,t0 ). Specifically, Figure 3 As shown in the figure, by adopting a distributed visual recognition algorithm, each camera is used to independently obtain the detection image and perform calculations to generate the relative position of the package in the independent coordinate system. Under the condition of ensuring the uniformity of the timestamp, the independent coordinate system of each camera is mapped to the virtual unified coordinate system through the visual recognition algorithm, thereby avoiding the high computational burden and error problems caused by the traditional image stitching method. Figure 2As shown in the figure, each camera performs independent target detection on the packages within its detection range. There is no need to stitch images from other cameras. Each camera independently runs the target detection algorithm through parallel computing, extracts the spatial features of the package through the convolutional neural network, identifies the location of the package and generates coordinates. Distributed computing breaks down the tasks to be completed independently by each camera, reducing the high computing burden brought by traditional centralized computing or image stitching. The use of convolutional neural networks to achieve accurate target detection avoids the image overlap and misalignment problems that may be caused by traditional stitching methods.

[0047] The present invention is further configured to generate the absolute coordinates of all detection targets by performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, including:

[0048] Set the coordinate values ​​of the first group of cameras on the left to the absolute coordinate values ​​in the unified coordinate system;

[0049] For cameras other than the first group of cameras on the left, compensate the target coordinate values ​​in their respective coordinate systems and convert them into absolute coordinates in the unified coordinate system required by the separation algorithm. The compensated coordinates are (X i+(i-1)Δd,t0 , Y j,t0 ), where Δd is the relative displacement between cameras.

[0050] Specifically, Figure 4 As shown in the figure, the problem of unifying independent coordinate systems in a multi-camera system is solved by the compensatory coordinate transformation method. Specifically, by setting the first group of cameras on the left as the benchmark of the unified coordinate system and generating compensation values ​​based on relative positions for other cameras, the independent coordinates of each camera are mapped to the unified coordinate system, thereby achieving the unification of the absolute coordinates of all detection targets in the entire system; the detection range of the first group of cameras on the left is the starting point of the entire system, so the target coordinates detected by it are directly used as the absolute coordinates of the unified coordinate system without any compensatory transformation. For other cameras, there is a relative position offset between their independent coordinate systems and the unified coordinate system. The relative offset of each camera is the distance between it and the detection range of the first group of cameras on the left. The compensatory coordinate transformation ensures that the detection results of all cameras can be described in the same coordinate system, avoiding conflicts between independent coordinate systems, and finally running the separation strategy algorithm based on the unified coordinate system.

[0051] The present invention is further configured to effectively filter the absolute coordinates of all the detected targets through a filter to eliminate duplicate coordinates, including:

[0052] The width of the camera boundary area is set to be larger than the maximum size of the package in the system to ensure that the package is effectively detected by only one camera in the boundary area. The camera on the left side of the boundary area is set as camera 1, and the camera on the right side of the boundary area is set as camera 2.

[0053] When the package tail coordinate X ib >X1, the coordinates of camera 1 are marked as invalid and not uploaded to the separation strategy algorithm module, while the coordinates of camera 2 are marked as valid and start to be uploaded;

[0054] When the package tail coordinate X ib ≤X1, the coordinates of the second camera are marked as invalid and not uploaded to the separation strategy algorithm module. ib is the horizontal coordinate of the tail of the camera No. 1 package, X1 is the horizontal coordinate of the boundary point of the intersection area between camera No. 1 and camera No. 2. If the detection area of ​​camera No. 1 is from X start To X end , and the width of the boundary area between camera 1 and camera 2 is L, X1=X end -L. Specific, such as Figure 5 As shown in the figure, by setting a filter in the camera boundary area, the problem of duplicate coordinates caused by overlapping fields of view in multi-camera detection is eliminated. Furthermore, by setting the width of the boundary area larger than the maximum size of the system package and adopting a coordinate switching mechanism, it is ensured that only one camera provides effective detection coordinates for the target in the boundary area, thereby achieving the accuracy of package detection and the stability of the separation strategy; through the filter and coordinate switching mechanism, the duplicate detection coordinates in the boundary area are effectively filtered to ensure the uniqueness of package detection.

[0055] The present invention is further configured such that the size of the single-piece separation module is a standardized design, the length and width of the module are set to 600mm and 150mm, and the splicing or expansion of multiple modules is supported according to different needs, and the camera is arranged directly above the single-piece separation module. Specifically, a modular system design is achieved by setting a single-piece separation module of standardized size and installing the camera directly above the module. The standardized module size supports flexible splicing and expansion to meet the needs of different logistics scenarios, while providing a precise visual detection basis through the camera viewing angle directly above. This design reduces system complexity, improves deployment flexibility and detection accuracy, and the standardized design provides a unified benchmark for module splicing and expansion.

[0056] The present invention is further configured such that the separation strategy algorithm module calculates the separation path of each package in real time according to the absolute coordinates of all detection targets, and guides the package to the target separation outlet through the actuator of the control system. Specifically, through the setting of the separation strategy algorithm module, the separation path of the package is calculated in real time according to the absolute coordinates of all detection targets, and the package is accurately guided to the target separation outlet by controlling the actuator. This design realizes the automation and intelligence of the whole process from coordinate detection to path planning, and then to the execution of separation operations, ensuring the efficiency and accuracy of package separation. The planning of the separation path needs to comprehensively consider the following factors: the current position of the package: that is, the absolute coordinates of the package; the target outlet position: the separation outlet coordinates preset in the system; the conveyor belt speed: used to dynamically adjust the path and control the timing; physical obstacles: to avoid collision or interference with other packages during the path planning process. The actuator completes the physical guidance of the package according to the path control instructions output by the separation strategy algorithm module. The above contents can all be achieved through existing technologies and will not be elaborated here. The separation strategy algorithm ensures that each package can accurately reach the target exit by calculating the path in real time, which greatly reduces the separation error rate. The real-time path planning and control instruction generation enable the system to efficiently handle high-speed moving packages and adapt to high-throughput logistics scenarios.

[0057] The present invention is further configured to add a camera to the human-machine system after setting the system configuration parameters, specify the detection range, and realize the modular increase or decrease of the single-piece separation module and the camera, so that the package can be separated quickly and efficiently. Specifically, by adding a human-machine interaction module to the system configuration parameters, the user can flexibly configure the camera and the detection range, thereby realizing the modular increase or decrease of the number of single-piece separation modules and cameras. This design improves the scalability and adaptability of the system, and can quickly adjust the system scale to meet the needs of different logistics scenarios, while ensuring the rapid and efficient separation of packages.

[0058] The present invention is further configured such that the method can adapt to the detection requirements of packages of different sizes by dynamically adjusting the width and boundary coordinates of the camera junction area. Specifically, by dynamically adjusting the width and boundary coordinates of the camera junction area, flexible adaptation to packages of different sizes is achieved. The junction area is the overlapping part of the detection range in a multi-camera system. The dynamic adjustment of its width and boundary can ensure the detection accuracy of the package in the junction area and avoid the problem of missed detection or repeated detection. Through this method, the package detection requirements in various logistics scenarios can be efficiently met.

[0059] The present invention is further configured such that the target visual processing based on the convolutional neural network includes a dense package visual real-time accurate detection algorithm based on the YOLOv5n network. Specifically, by introducing a target visual processing algorithm based on a convolutional neural network (CNN), a lightweight YOLOv5n network is used to perform real-time accurate detection of dense packages in logistics scenarios. With its efficient detection speed and small model size, YOLOv5n adapts to the characteristics of a large number of packages and high detection speed requirements in logistics scenarios, while ensuring detection accuracy, and providing high-quality coordinate data for subsequent package separation strategies.

[0060] After actual testing of multiple projects, the length of the test equipment ranged from 2.1 meters to 8.7 meters, and the detection targets ranged from medicine boxes (minimum length 7 cm) to oversized packages (length over 1 meter). Under this sample, the patented method can save 30% of the deployment labor cost, increase the deployment time by 50%, and reduce the training time by 60%. At the same time, the system detection accuracy and separation efficiency were improved, with the efficiency increased by 4.5%, greatly improving the flexibility of the system.

[0061] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0062] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0063] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0064] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0065] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0066] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0067] In the several embodiments provided in the present application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0068] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0069] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0070] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0071] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for designing a modular distributed system for high-speed visual package separation, wherein the system is provided with a single-piece separation module of uniform size, characterized in that: The method comprises: Set the row and column distribution and size values ​​of the single separation module according to the system configuration parameters; According to the detection range of each camera, an independent coordinate system corresponding to each camera is generated, and the same visual recognition algorithm is entered to generate the absolute position coordinates of the detection target of each camera; By performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, the absolute coordinates of all detection targets are generated; Effectively filtering the absolute coordinates of all the detected targets through a filter to eliminate duplicate coordinates; The absolute coordinates of all detected targets after eliminating duplicate coordinates are input into the separation strategy algorithm module of the system to achieve single-piece separation of packages.

2. According to claim 1, a method for designing a modular distributed system for high-speed package separation, characterized in that: The generation of the independent coordinate system is based on the detection range of each camera and is automatically generated by inputting the row and column distribution of the single separation module and the size parameters of each module.

3. According to claim 1, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: The visual recognition algorithm uses each camera to obtain the detection image independently, and through the target visual processing based on the convolutional neural network, the relative position of the package in the independent coordinate system is calculated independently in a distributed manner. Under the condition of ensuring the uniformity of the timestamp, the virtual unified coordinate system is realized through the visual recognition algorithm to avoid obtaining the unified coordinate system by splicing images. Among them, at the same time t0, the real real position (X i,t0 , Y j,t0 ).

4. According to claim 3, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: By performing a unified compensation coordinate transformation on the absolute position coordinates of the detection target of each camera, the absolute coordinates of all detection targets are generated, including: Set the coordinate values ​​of the first group of cameras on the left to the absolute coordinate values ​​in the unified coordinate system; For cameras other than the first group of cameras on the left, compensate the target coordinate values ​​in their respective coordinate systems and convert them into absolute coordinates in the unified coordinate system required by the separation algorithm. The compensated coordinates are (X i+(i-1)Δd,t0 , Y j,t0 ), where Δd is the relative displacement between cameras.

5. According to claim 1, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: The absolute coordinates of all the detected targets are effectively filtered through filters to eliminate duplicate coordinates, including: The width of the camera boundary area is set to be larger than the maximum size of the package in the system to ensure that the package is effectively detected by only one camera in the boundary area. The camera on the left side of the boundary area is set as camera 1, and the camera on the right side of the boundary area is set as camera 2. When the package tail coordinate X ib >X1, the coordinates of camera 1 are marked as invalid and not uploaded to the separation strategy algorithm module, while the coordinates of camera 2 are marked as valid and start to be uploaded; When the package tail coordinate X ib ≤X1, the coordinates of the second camera are marked as invalid and not uploaded to the separation strategy algorithm module. ib is the horizontal coordinate of the tail of the camera No. 1 package, X1 is the horizontal coordinate of the boundary point of the intersection area between camera No. 1 and camera No.

2. If the detection area of ​​camera No. 1 is from X start To X end , and the width of the boundary area between camera 1 and camera 2 is L, X1=X end -L.

6. According to claim 1, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: The size of the single-piece separation module is a standardized design, and the length and width of the module are set to 600mm and 150mm. It supports splicing or expansion of multiple modules according to different needs, and the camera is set directly above the single-piece separation module.

7. According to claim 1, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: The separation strategy algorithm module calculates the separation path of each package in real time according to the absolute coordinates of all detection targets, and guides the package to the target separation outlet through the actuator of the control system.

8. According to claim 1, a method for designing a modular distributed system for high-speed visual package separation is characterized in that: After setting the system configuration parameters, add cameras to the human-machine system, specify the detection range, and realize the modular increase or decrease of single-piece separation modules and cameras, so that packages can be separated quickly and efficiently.

9. The method for designing a modular distributed system for high-speed visual package separation according to claim 1, characterized in that: The method can adapt to the detection requirements of packages of different sizes by dynamically adjusting the width and boundary coordinates of the camera boundary area.

10. The method for designing a modular distributed system for high-speed visual package separation according to claim 2, characterized in that: Target vision processing based on convolutional neural networks includes a dense parcel vision real-time accurate detection algorithm based on the YOLOv5n network.

Images