Method and device for detecting the movement of goods

Abstract

Embodiments of this specification provide a method and apparatus for detecting the movement of goods. In this method, point cloud data within the field of view collected by a radar device is obtained, and the radar device is used to monitor the storage goods within the field of view in real time; the adjacent two frames of point cloud data obtained are compared to determine whether there is goods movement; when it is determined that there is goods movement, two frames of images corresponding to the two frames of point cloud data are obtained, where the images are collected by a camera device through real-time shooting of the storage goods, and the acquisition times of the two frames of point cloud data and the corresponding two frames of images are the same; and the moving goods in the storage goods are determined according to the two frames of images.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of Internet of Things technology. Specifically, they relate to a method and device for detecting the movement of goods. Background Art

[0002] In the warehousing and logistics industry, goods are stored and managed centrally. With the development and wide application of technology, technologies such as artificial intelligence and the Internet of Things are applied to the field of goods safety monitoring. The way of goods safety monitoring has been upgraded from traditional manual guarding to using technical means for goods monitoring.

[0003] Currently, the movement monitoring of goods mainly adopts a combination of vision and artificial intelligence. In this method, camera devices are deployed in the environment where the goods are located. The camera devices collect images of the goods in real time and transmit the collected images to a server device with relatively high computing power in real time. The server device uses AI-based technical means to analyze and process each frame of the image to determine whether the goods have moved. Compared with traditional manual guarding, monitoring goods through AI-based technical means not only saves manpower but also has higher accuracy in detecting the movement of goods. Summary of the Invention

[0004] In view of the above, the embodiments of this specification provide a method and device for detecting the movement of goods. Through the embodiments of this specification, the point cloud data generated by a radar device is used to monitor whether there is any movement of goods, improving the accuracy of detecting the movement of goods. In addition, only when the goods move, the frame images of the goods are further used to determine the moving goods, avoiding the problem of processing continuous frame images, thereby reducing the computing power requirements. Lower computing power reduces the dependence on high requirements for hardware configuration and saves computing resources.

[0005] According to one aspect of the embodiments of this specification, a method for detecting the movement of goods is provided, including: obtaining point cloud data within the field of view collected by a radar device, where the radar device is used to monitor the warehousing goods within the field of view in real time; comparing two adjacent frames of the obtained point cloud data to determine whether there is any movement of goods; when it is determined that there is movement of goods, obtaining two frames of images corresponding to the two frames of point cloud data, where the images are collected by a camera device through real-time photographing of the warehousing goods, and the acquisition times of the two frames of point cloud data and the corresponding two frames of images are the same; and determining the moving goods in the warehousing goods according to the two frames of images.

[0006] According to another aspect of the embodiments of the present specification, there is also provided a method for detecting the movement of goods, which is executed by an edge server. The edge server is communicatively connected to a radar device and a camera device respectively, and is also communicatively connected to a cloud server. The radar device is used for real-time monitoring of storage goods within the field of view, and the camera device is used for real-time photographing of the storage goods. The method includes: obtaining the point cloud data within the field of view collected by the radar device; comparing two adjacent frames of the obtained point cloud data to determine whether there is goods movement; when it is determined that there is goods movement, obtaining two frames of images corresponding to the two frames of point cloud data, and the acquisition times of the two frames of cloud data and the corresponding two frames of images are the same; and invoking the visual detection algorithm configured by the cloud server to determine the moving goods in the storage goods according to the two frames of images.

[0007] According to another aspect of the embodiments of the present specification, there is also provided a device for detecting the movement of goods, which is applied to an edge server. The edge server is communicatively connected to a radar device and a camera device respectively, and is also communicatively connected to a cloud server. The radar device is used for real-time monitoring of storage goods within the field of view, and the camera device is used for real-time photographing of the storage goods. The device includes: a point cloud data acquisition unit configured to obtain the point cloud data within the field of view collected by the radar device; a point cloud data comparison unit configured to compare two adjacent frames of the obtained point cloud data to determine whether there is goods movement, and when it is determined that there is goods movement, trigger the image acquisition unit; the image acquisition unit configured to obtain two frames of images corresponding to the two frames of point cloud data, and the acquisition times of the two frames of cloud data and the corresponding two frames of images are the same; and an algorithm invocation unit configured to invoke the visual detection algorithm configured by the cloud server to determine the moving goods in the storage goods according to the two frames of images.

[0008] According to another aspect of the embodiments of the present specification, there is also provided an electronic device, including: at least one processor, a memory coupled to the at least one processor, and a computer program stored on the memory. The at least one processor executes the computer program to implement the method for detecting the movement of goods as described in any one of the above.

[0009] According to another aspect of the embodiments of the present specification, there is also provided a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method for detecting the movement of goods as described above.

[0010] According to another aspect of the embodiments of the present specification, there is also provided a computer program product, including a computer program. When the computer program is executed by a processor, it implements the method for detecting the movement of goods as described in any one of the above. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] By referring to the following drawings, a further understanding of the essence and advantages of the content of the embodiments of this specification can be achieved. In the drawings, similar components or features may have the same reference numerals.

[0012] Figure 1 FIG. shows a schematic diagram of an example of the hardware device deployment structure for implementing the goods movement detection method according to an embodiment of this specification.

[0013] Figure 2 FIG. shows a schematic diagram of another example of the hardware device deployment structure for implementing the goods movement detection method according to an embodiment of this specification.

[0014] Figure 3 FIG. shows a flowchart of an example of the method for detecting goods movement according to an embodiment of this specification.

[0015] Figure 4 FIG. shows a flowchart of an example of comparing two adjacent frames of point cloud data according to an embodiment of this specification.

[0016] Figure 5 FIG. shows a schematic diagram of an example of spatial partitioning using the octree algorithm according to an embodiment of this specification.

[0017] Figure 6 FIG. shows a flowchart of another example of comparing two adjacent frames of point cloud data according to an embodiment of this specification.

[0018] Figure 7 FIG. shows a flowchart of another example of the method for detecting goods movement according to an embodiment of this specification.

[0019] Figure 8 FIG. shows a block diagram of an example of the apparatus for detecting goods movement according to an embodiment of this specification.

[0020] Figure 9 FIG. shows a block diagram of an example of the electronic device for implementing the goods movement detection method according to an embodiment of this specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] The subject matter described herein will be discussed with reference to example embodiments below. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and is not a limitation on the scope of protection, applicability, or examples set forth in the claims. The functions and arrangements of the elements discussed can be changed without departing from the scope of protection of the content of the embodiments of this specification. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0022] As used herein, the term "comprising" and its variants denote open terms meaning "including but not limited to". The term "based on" means "at least partially based on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first", "second", etc. can refer to different or the same objects. Other definitions, whether explicit or implicit, may be included below. Unless explicitly specified in the context, the definition of a term is consistent throughout the specification.

[0023] In the warehousing and logistics industry, goods are stored and managed centrally. With the development and wide application of technology, technologies such as artificial intelligence and the Internet of Things are applied to the field of goods safety monitoring, and the way of goods safety monitoring has been transformed and upgraded from traditional manual guarding to using technical means for goods monitoring.

[0024] Currently, the mobile monitoring of goods mainly adopts a combination of vision and artificial intelligence. In this method, camera devices are deployed in the environment where the goods are located. The camera devices collect images of the goods in real time and transmit the collected images to a server device with relatively high computing power in real time. The server device uses AI-based technical means to analyze and process each frame of the image to determine whether the goods have moved. Compared with traditional manual guarding, monitoring goods through AI-based technical means not only saves manpower but also has higher accuracy in detecting the movement of goods.

[0025] However, the camera devices collect images in real time and transmit the collected images as a video stream to the server device. The server device needs to process the continuous video stream, which results in the server device requiring very high computing power. Moreover, if the image processing is performed locally on the server device, there are also very high requirements for the hardware configuration of the server device.

[0026] In view of the above, the embodiments of the present specification provide a method and device for detecting the movement of goods. In this method, point cloud data within the field of view collected by a radar device is obtained, and the adjacent two frames of point cloud data obtained are compared to determine whether there is goods movement; when it is determined that there is goods movement, two frames of images corresponding to the two frames of point cloud data are obtained, where the images are collected by a camera device through real-time shooting of the storage goods, and the acquisition times of the two frames of point cloud data and the corresponding two frames of images are the same; and the moving goods in the storage goods are determined according to the two frames of images. Through the embodiments of the present specification, whether there is goods movement is monitored through the point cloud data generated by the radar device, improving the accuracy of goods movement detection. In addition, only when there is goods movement, the moving goods are further determined according to the frame images of the goods, avoiding the problem of processing continuous frame images, thereby reducing the requirement for computing power. The lower computing power reduces the dependence on high requirements for hardware configuration and saves computing resources.

[0027] The following describes in detail the method and device for detecting the movement of goods provided by the embodiments of the present specification with reference to the accompanying drawings.

[0028] Figure 1 FIG. 7 shows a schematic diagram of an example 100 of the hardware device deployment structure for implementing the goods movement detection method according to the embodiments of the present specification.

[0029] As Figure 1 shown, the hardware device for implementing the goods movement detection method may include a server device, a camera device, and a radar device. The server device is communicatively connected to the radar device and the camera device through a network respectively.

[0030] In the embodiments of the present specification, the radar device is used to monitor the storage goods within the field of view and can sense and output point cloud data. The radar device may include a lidar, a millimeter-wave radar, etc. For example, the lidar may include LiDAR, etc., and the millimeter-wave radar may include 4D mmRadar, etc.

[0031] The field of view of the radar device can be set according to the position of the storage goods so that the storage goods are included in the space of the field of view of the radar device. The radar device can detect the space within the field of view by emitting radar waves to determine the spatial positions of various objects in the space within the field of view, and the spatial positions can be characterized by point cloud data.

[0032] The camera device can perform real-time shooting on the storage goods to collect real-time images of the storage goods. The camera device and the radar device can be synchronized, and each frame of image collected by the camera device corresponds one-to-one to each frame of point cloud data collected by the radar device, that is, the acquisition time of each frame of image is the same as the acquisition time of the corresponding frame of point cloud data.

[0033] The server device can obtain corresponding point cloud data and images from the radar device and the camera device, and provide computing power to perform corresponding processing on the point cloud data and images. For example, it can determine whether there is cargo movement based on two adjacent frames of point cloud data, and use a visual detection algorithm to calculate two frames of images to determine the moving cargo. In this example, the server device can perform corresponding operations locally.

[0034] In some embodiments, the network can be any one or more of a wired network or a wireless network. Examples of the network can include but are not limited to cable networks, fiber optic networks, telecommunications networks, enterprise internal networks, the Internet, local area networks (LANs), wide area networks (WANs), wireless local area networks (WLANs), metropolitan area networks (MANs), public switched telephone networks (PSTNs), Bluetooth networks, ZigBee networks, near field communication (NFC), in-device buses, in-device lines, etc. or any combination thereof.

[0035] Figure 2 FIG. shows a schematic diagram of another example 200 of the hardware device deployment structure for implementing the cargo movement detection method according to the embodiments of the present specification.

[0036] As Figure 2 shown, based on the above Figure 1 shown example, the server device can communicate with the cloud server. The cloud server can provide various services such as data storage services and image processing services. Thus, the server device can call the services provided by the cloud server. For example, the server device can call the data storage service configured in the cloud server to store corresponding data (such as point cloud data and image data of the moving cargo) in the cloud. Also, for example, the image processing service provided by the cloud server can be a visual detection algorithm. Thus, the server device can call the visual detection algorithm configured in the cloud server to process the image data to determine the moving cargo. In one example, the server device can include an edge server.

[0037] In this example, by providing various services through the cloud server for the server device to call, without the need for the server device to execute locally, the load pressure on the server device and the requirement for computing power can be reduced. In addition, storing data in the cloud can save the local storage space of the server device.

[0038] Figure 3 FIG. shows a flowchart of an example 300 of the method for detecting cargo movement according to the embodiments of the present specification.

[0039] As Figure 3 shown, at 310, the point cloud data within the field of view collected by the radar device can be obtained.

[0040] In the embodiments of this specification, the radar device can collect point cloud data within its field of view in real time, store the collected point cloud data locally, and also send it to the server device. Thus, the server device can obtain point cloud data from the radar device and also from local storage, and the obtained point cloud data can be used to perform subsequent comparison operations.

[0041] The acquisition of point cloud data can be frame-based. Each time the acquired point cloud data can include one or more frames, and each frame of point cloud data can include multiple points. Each frame of point cloud data can be used to represent the spatial positions of objects within the field of view of the radar device at the moment corresponding to that frame.

[0042] In one example, the acquisition of point cloud data can be performed at a specified time interval, that is, the time interval between two adjacent frames of acquired point cloud data is a specified duration. The specified duration can be set customarily, for example, 1 second or 10 seconds. In this example, the point cloud data used for comparison can be part of the point cloud data collected by the radar device. By acquiring at intervals, the amount of data of the point cloud data for performing comparison operations can be reduced, thereby reducing the load pressure and required computing power of the server device.

[0043] In one acquisition method, when the radar device stores the collected point cloud data locally, the server device can obtain frame point cloud data from the radar device at a time interval of the specified duration according to the frame time corresponding to each frame of point cloud data, and the frame time interval between two adjacent frames of acquired point cloud data is the specified duration. The frame time corresponding to each frame of point cloud data can be the generation time of that frame of point cloud data.

[0044] In another acquisition method, when the radar device sends all the collected point cloud data to the server device, the server device can store all the point cloud data in the local memory, so that the server device can obtain frame point cloud data from the local memory at a time interval of the specified duration according to the frame time corresponding to each frame of point cloud data.

[0045] After the server device obtains each frame of point cloud data, it can store the obtained each frame of point cloud data in the local cache. In one storage method, the local cache can divide a specified first area for storing two adjacent frames of point cloud data. The two adjacent frames of point cloud data cached in this first area are the two frames of point cloud data to be subjected to a comparison operation. When the previous frame of point cloud data in the two adjacent frames of point cloud data stored in the first area is taken out for performing the comparison operation, at this time, the latter frame of point cloud data in the two adjacent frames of point cloud data in the first area can become the previous frame of point cloud data. Then, a latter frame of point cloud data that is separated from this previous frame of point cloud data by a specified time duration can be obtained and stored in the first area as the latter frame of point cloud data in the two adjacent frames of point cloud data stored in the first area.

[0046] In one example, the first area can be divided into two sub-areas: a first sub-area and a second sub-area. The first sub-area is used to store the previous frame of point cloud data in the two adjacent frames of point cloud data, and the second sub-area is used to store the latter frame of point cloud data in the two adjacent frames of point cloud data. The point cloud data stored in the first sub-area will be preferentially used for performing the comparison operation. After the point cloud data in the first sub-area is taken out for performing the comparison operation, the point cloud data in the second sub-area can be transferred to the first sub-area for storage, so that this point cloud data becomes the previous frame of point cloud data in the two adjacent frames of point cloud data. Then, a latter frame of point cloud data that is separated from this previous frame of point cloud data by a specified time duration can be obtained and stored in the second sub-area as the latter frame of point cloud data in the two adjacent frames of point cloud data stored in the first area.

[0047] At 320, the two adjacent frames of point cloud data obtained can be compared.

[0048] At 330, it can be determined whether there is cargo movement according to the comparison result at 320. If there is cargo movement, the operation at 340 is performed; if there is no cargo movement, the operation at 310 is returned, that is, continue to obtain two other adjacent frames of point cloud data to monitor whether there is cargo movement.

[0049] The two adjacent frames of point cloud data can respectively represent the spatial positions of the storage cargo within the field of view at the two time points corresponding to these two frames of point cloud data, then the two adjacent frames of point cloud data can be compared one by one. When the two adjacent frames of point cloud data are the same, it can be determined that there is no cargo movement; when there are differences in the point cloud data compared one by one in the two adjacent frames of point cloud data, it can be determined that there is cargo movement.

[0050] Figure 4 A flowchart of an example 400 for comparing two adjacent frames of point cloud data according to an embodiment of the present specification is shown.

[0051] As Figure 4As shown in 321-1, the obtained adjacent two-frame point cloud data can be respectively substituted into the pre-divided first field-of-view division space to obtain the tree structure indexes corresponding to the two-frame point cloud data.

[0052] In this example, the first field-of-view division space is obtained by dividing the space within the field of view of the radar device. The first field-of-view division space can include a plurality of divided small three-dimensional spaces, and these plurality of small three-dimensional spaces can constitute the space within the complete field of view.

[0053] The first field-of-view division space can be obtained during the preprocessing process, and the preprocessing process can be executed before implementing the technical solution of this specification. During the preprocessing process, the octree algorithm can be used to divide the space within the field of view according to the point cloud data generated by the radar device to obtain the first field-of-view division space.

[0054] In one example, the three-dimensional space within the field of view of the radar device can be used as a three-dimensional space to be divided. The radar device can emit radar waves into the space within the field of view to obtain the point cloud in the space within the field of view. The number of radar waves emitted by the radar device each time is the same, so the number of point clouds obtained each time is the same, and further the number of point clouds in each frame of point cloud data is the same.

[0055] The octree algorithm can be used to divide the three-dimensional space within the field of view into 8 quadrants to obtain 8 three-dimensional spaces to be divided, and each three-dimensional space to be divided can include the point cloud generated by the radar device. The number of point clouds included in each three-dimensional space to be divided can be counted. For each three-dimensional space to be divided, it can be determined whether the number of point clouds included in the three-dimensional space to be divided is greater than the first quantity threshold. The first quantity threshold can be customized. For example, the first quantity threshold can be 128.

[0056] If the number of point clouds included in the three-dimensional space to be divided is greater than the first quantity threshold, then the three-dimensional space to be divided can be divided into 8 quadrants to further obtain 8 three-dimensional spaces to be divided. If the number of point clouds included in the three-dimensional space to be divided is not greater than the first quantity threshold, then the three-dimensional space to be divided will not be further divided and remains unchanged. The above method is used to recursively divide each three-dimensional space until the number of point clouds included in each obtained three-dimensional space is less than the first quantity threshold. After completing the space division, the space within the field of view can be divided into several small three-dimensional spaces, and the space sizes of each three-dimensional space can be different.

[0057] Figure 5 Shows a schematic diagram of an example of using the octree algorithm for space division according to an embodiment of this specification. As Figure 5As shown, the space within the field of view can be approximated as a cube. After the first spatial division, the cube can be divided into eight small cubes. Then, compare whether the number of point clouds included in each small cube is greater than the first quantity threshold. When the number of point clouds is greater than the first quantity threshold (for example, Figure 5 the upper-right small cube shown), this small cube can be further divided into eight even smaller cubes.

[0058] After the spatial division of the space within the field of view is completed, the space within the field of view can be divided into several small three-dimensional spaces. In addition, according to the process of spatial division, a tree-like structure index can be correspondingly obtained. During the spatial division process, the space within the field of view can be represented by a root node. Thereafter, each three-dimensional space to be divided can be represented by a leaf node. When a three-dimensional space is divided into eight small three-dimensional spaces, the root node or leaf node corresponding to this three-dimensional space and the leaf nodes corresponding to the eight small three-dimensional spaces have a parent-child relationship, and the two nodes with the parent-child relationship are mutually associated.

[0059] Take Figure 5 as an example. The space within the field of view is the largest three-dimensional space to be divided and is represented by a root node. When the space within the field of view is divided into eight small three-dimensional spaces, each corresponding small three-dimensional space can be represented by eight leaf nodes. The eight leaf nodes have a parent-child relationship with the root node, so the eight leaf nodes can be connected to the root node to represent the association between the two. For subsequent division of each three-dimensional space, the leaf nodes and the relationships between the leaf nodes can be constructed in this way. After the spatial division is completed, the tree-like structure formed by the relationships between the obtained root node and each leaf node is the tree-like structure index corresponding to the spatial division of the space within the field of view.

[0060] Returning to Figure 4 , after there is a pre-divided first field-of-view division space, the two adjacent frames of point cloud data obtained can be respectively substituted into the first field-of-view division space. After each frame of point cloud data is substituted into the first field-of-view division space, according to the position of the point cloud represented by each point cloud data in space, it can be determined which three-dimensional space in the first field-of-view division space the point cloud belongs to. Based on this, the attribution relationship between each point cloud in the frame of point cloud data and each three-dimensional space in the first field-of-view division space can be determined.

[0061] Since the spatial structure of the first field-of-view division space remains unchanged, the tree-like structures of the tree-like structure indexes corresponding to each frame of point cloud data are the same, and the tree-like structure of the tree-like structure index corresponding to each frame of point cloud data is the tree-like structure corresponding to the first field-of-view division space.

[0062] For each frame of point cloud data, according to the attribution relationship between each point cloud and each three-dimensional space in the first field-of-view division space, the number of point clouds included in each three-dimensional space in the first field-of-view division space and the point cloud data corresponding to each point cloud can be determined. Thus, in the obtained tree structure index, the number of point clouds included in each node and the point cloud data corresponding to each point cloud are the number of point clouds included in the three-dimensional space corresponding to the node and the point cloud data corresponding to each point cloud.

[0063] In 321-3, it is possible to determine whether there is cargo movement according to the tree structure indexes corresponding to two frames of point cloud data.

[0064] In the embodiments of this specification, it is determined whether there is cargo movement according to whether two tree structure indexes are the same or the degree of difference. The degree of difference can be measured by the number of different leaf nodes in the two tree structure indexes. The more the number of different leaf nodes, the greater the degree of difference; the fewer the number of different leaf nodes, the smaller the degree of difference. The cargo that moves can be all or part of the cargo in the warehouse.

[0065] In one example, the leaf nodes at the end of each branch in the tree structure indexes corresponding to two frames of point cloud data can be respectively compared correspondingly to count the number of different leaf nodes in the two tree structure indexes.

[0066] There are multiple leaf nodes in the tree structure index, and only the leaf nodes at the end of each branch in the tree structure index need to be compared. For Figure 5 example, the 8 leaf nodes associated with the root node are respectively leaf node 1, 2, 3, 4, 5, 6, 7, and 8. Among them, leaf nodes 3 and 7 are further associated with another 8 leaf nodes, which are: leaf nodes 3-1, 3-2, 3-3, 3-4, 3-5, 3-6, 3-7, and 3-8, and leaf nodes 7-1, 7-2, 7-3, 7-4, 7-5, 7-6, 7-7, and 7-8. Thus, the leaf nodes at the end of the branches of leaf nodes 3 and 7 are leaf nodes 3-1, 3-2, 3-3, 3-4, 3-5, 3-6, 3-7, and 3-8, and leaf nodes 7-1, 7-2, 7-3, 7-4, 7-5, 7-6, 7-7, and 7-8. Then, for this tree structure index, the leaf nodes to be compared include leaf nodes 1, 2, 4, 5, 6, and 8, leaf nodes 3-1, 3-2, 3-3, 3-4, 3-5, 3-6, 3-7, and 3-8, and leaf nodes 7-1, 7-2, 7-3, 7-4, 7-5, 7-6, 7-7, and 7-8.

[0067] When comparing two tree - like structure indexes, two leaf nodes in the two tree - like structure indexes that represent the same three - dimensional space can be compared. In one example, for the two leaf nodes being compared, the point cloud data included in the three - dimensional space represented by the two leaf nodes can be compared. The comparison content includes the number of point clouds represented by the point cloud data included in the two leaf nodes and the point cloud data of each point cloud. When the number of point clouds is different, the difference in the number of point clouds can be used as the number of different point cloud data. In addition, except for the same point cloud data included in the two leaf nodes, other different point cloud data can be used as the number of different point cloud data to be counted.

[0068] In this example, for the two nodes being compared, when the number of different point cloud data is greater than the third quantity threshold, it can be determined that the two leaf nodes are different; when the number of different point cloud data is not greater than the third quantity threshold, it can be determined that the two leaf nodes are the same. In another example, when the number of point clouds included in the two leaf nodes and / or the point cloud data corresponding to each point cloud are different, it can be determined that the two leaf nodes are different.

[0069] After counting the number of different leaf nodes in the two tree - like structure indexes, it can be determined whether there is cargo movement based on the number of different leaf nodes. Specifically, when the number of different leaf nodes is greater than the second quantity threshold, it can be determined that there is cargo movement; when the number of different leaf nodes is not greater than the second quantity threshold, it can be determined that there is no cargo movement.

[0070] In one example, among the leaf nodes being compared, when there are different leaf nodes, it can be determined that the two tree - like structure indexes are different, and thus it can be determined that there is cargo movement. When the leaf nodes being compared are all the same, it can be determined that the two tree - like structure indexes are the same, and thus it can be determined that there is no cargo movement.

[0071] Figure 6 The flowchart of another example 600 for comparing adjacent two - frame point cloud data according to an embodiment of this specification is shown.

[0072] As Figure 6 shown in 322 - 1, the octree algorithm can be used to perform spatial partitioning on the space within the field of view according to one of the two - frame point cloud data to generate a second field - of - view partitioned space and obtain the tree - like structure index corresponding to this frame of point cloud data.

[0073] In this example, one of the two - frame point cloud data used as the basis for spatial partitioning can be any one of the two - frame point cloud data. For example, this one - frame point cloud data is the previous - frame point cloud data of the two - frame point cloud data. The process of spatial partitioning can refer to the relevant operation descriptions for the first field - of - view partitioned space above Figure 5 and will not be elaborated here.

[0074] In this example, the second field of view division space is obtained by dividing the space within the field of view of the radar device. The second field of view division space may include a plurality of divided small three-dimensional spaces, and these plurality of small three-dimensional spaces may constitute the space within the complete field of view. The structure of the three-dimensional spaces distributed in the second field of view division space may be the same as or different from the structure of the three-dimensional spaces distributed in the first field of view division space.

[0075] The second field of view division space may only include the composition information of each three-dimensional space after the division of the space within the field of view, and does not include the frame point cloud data information based on which the space is divided. The tree structure index corresponding to one of the frames of point cloud data obtained may include the frame point cloud data information, and can be used to represent the distribution of the frame point cloud data in the second field of view division space.

[0076] In 322-3, another frame of point cloud data in the two frames of point cloud data can be substituted into the second field of view division space to obtain the tree structure index corresponding to the other frame of point cloud data.

[0077] In this example, the obtained tree structure index can be used to represent the distribution of another frame of point cloud data in the second field of view division space. The two tree structure indexes corresponding to the two frames of point cloud data are both generated based on the second field of view division space, so the tree structures of the two generated tree structure indexes are the same.

[0078] The operation of 322-3 can refer to the operation description of the above Figure 5 example, which will not be elaborated here.

[0079] In 322-5, it is determined whether there is cargo movement according to the tree structure indexes corresponding to the two frames of point cloud data.

[0080] In one example, the leaf nodes at the end of each branch in the tree structure indexes corresponding to the two frames of point cloud data are respectively compared correspondingly to count the number of different leaf nodes in the two tree structure indexes; and when the number of different leaf nodes is greater than the second quantity threshold, it is determined that there is cargo movement.

[0081] In one example, for two leaf nodes in the tree structure indexes corresponding to the two frames of point cloud data that are used to represent the same three-dimensional space, the point cloud data included in the three-dimensional spaces represented by the two leaf nodes are compared; when the number of different point cloud data is greater than the third quantity threshold, it can be determined that the two leaf nodes are different; and when the number of different point cloud data is not greater than the third quantity threshold, it can be determined that the two leaf nodes are the same.

[0082] The operation of 322-5 can refer to the relevant operation description of the above Figure 5 example, which will not be elaborated here.

[0083] In one example, before comparing two adjacent frames of collected point cloud data, the two adjacent frames of collected point cloud data can be filtered to remove the outlier point cloud data introduced by noise. The filtering methods can include at least one of Statistical Outlier Removal, Conditional Removal, and Radius Outlier Removal.

[0084] By using the point cloud generated by the radar device provided in the embodiments of this specification to monitor whether the goods move, the position of the goods in space can be more accurately located by using the point cloud data. Each point cloud data can represent the position of a part of the goods, and several point cloud data are used to represent the complete position of the goods. The goods as a whole are converted into being represented by several point cloud data. Even if there is a slight movement in a part of the goods, it will cause the change of the point cloud data of that part. Therefore, based on the comparison of the two adjacent frames of point cloud data, the movement of the goods can be more accurately detected.

[0085] Back to Figure 3 , when it is determined that the goods move, at 340, obtain two frames of images corresponding to the two frames of point cloud data.

[0086] In the embodiments of this specification, the images are collected by a camera device by taking real-time pictures of the storage goods. The camera device can collect images of the storage goods in real time. The camera device can store the collected images in a specified memory, and can also send them to the server device for storage.

[0087] In the embodiments of this specification, the collection times of the two frames of point cloud data and the corresponding two frames of images are the same. The two frames of images with the same collection time can be obtained according to the two frames of point cloud data. In one example, two frames of images with the same collection time as the two frames of point cloud data can be screened out from the images stored in the specified memory or locally on the server device.

[0088] The server device can screen out the corresponding two frames of images from the images collected by the camera device. At this time, the two frames of images are in a to-be-processed state. The screened two frames of images can be stored in the local cache. In this way, when it is necessary to obtain images for processing, they can be directly obtained from the cache. In one storage method, the local cache can divide a specified second area for storing the two frames of images corresponding to the two frames of point cloud data. The two frames of images cached in this second area are the two frames of images to be obtained and used to perform the operation of determining the moving goods. The generation times of the two frames of images are different, and the generation time of the previous frame of the two frames of images is earlier than that of the latter frame of images.

[0089] When the previous frame image among the two frame images stored in the second region is taken out for performing the operation of determining the moving goods, at this time, the subsequent frame image among the two frame images stored in the second region can become the previous frame image among the two frame images stored in the second region. Then, it is possible to continue to obtain the image corresponding to the subsequent frame point cloud image and store it as the subsequent frame image among the two frame images stored in the second region in the second region.

[0090] In one example, the second region can be divided into two sub-regions: a third sub-region and a fourth sub-region. The third sub-region is used to store the previous frame image among the two frame images, that is, the image corresponding to the frame point cloud data in the first sub-region. The fourth sub-region is used to store the subsequent frame image among the two frame images, that is, the image corresponding to the frame point cloud data in the second sub-region. The image stored in the third sub-region will be preferentially used for performing the operation of determining the moving goods. After the image in the third sub-region is taken out to perform the operation of determining the moving goods, the image in the fourth sub-region can be transferred to the third sub-region for storage, so that this image becomes the previous frame image among the two stored frame images. Then, it is possible to obtain the image corresponding to the frame point cloud data stored in the second sub-region and store it as the subsequent frame image among the two frame images stored in the second region in the fourth sub-region.

[0091] In one example, the second region and the first region for storing two adjacent frame point cloud data can be merged into one region, and this region can be divided into: a first sub-region for storing the previous adjacent frame point cloud data, a second sub-region for storing the subsequent adjacent frame point cloud data, a third sub-region for storing the frame image corresponding to the frame point cloud data in the first sub-region, and a fourth sub-region for storing the frame image corresponding to the frame point cloud data in the second sub-region.

[0092] At 350, the moving goods in the warehoused goods can be determined according to the two obtained frame images.

[0093] In one example, a visual detection algorithm can be used to process the two obtained frame images to determine the position difference of the warehoused goods in the two frame images, and the goods with different positions are the moving goods. In this example, the visual detection algorithm can be executed by a trained model, and the model can perform processing such as region selection, feature extraction, and classification regression on the input image.

[0094] The determination method of the moving goods can include outputting the feature information of the moving goods, marking the moved goods on the image output by the model, etc.

[0095] In one example, the information of two adjacent frames of cloud data, two frames of images, and moving goods can be uploaded to the blockchain to ensure the security of the stored data and prevent the stored data from being tampered with. The execution device for performing the uploading operation can be a server device, or the server device can call the cloud server to execute.

[0096] In this example, the information of two adjacent frames of cloud data, two frames of images, and moving goods can be stored in the cloud. Then, when performing the uploading operation on the information of two adjacent frames of cloud data, two frames of images, and moving goods, the hash values of the information of two adjacent frames of cloud data, two frames of images, and moving goods can be stored in the blockchain. In another example, the original data of the information of two adjacent frames of cloud data, two frames of images, and moving goods can also be uploaded to the blockchain for storage.

[0097] In an application scenario, a normal time period and an abnormal time period can be set. For example, the daytime can be set as the normal time period, and the night can be set as the abnormal time period. The storage goods can be monitored in real time during both the normal time period and the abnormal time period. When the time point when the goods are detected to move belongs to the specified abnormal time period, a warning can be issued to timely warn of the risks existing in the goods.

[0098] Through the technical solution provided by the embodiments of this specification, a radar is used to monitor whether there is any movement of goods. Only when the goods move, the goods that have moved are further determined based on the frame images of the goods. When no movement of goods is detected, the frame images are not processed, avoiding the problem of requiring high computing power to continuously process frame images. The lower computing power reduces the dependence on high hardware configuration requirements and saves computing resources.

[0099] Figure 7 The flowchart of another example 700 of the method for detecting the movement of goods according to the embodiments of this specification is shown.

[0100] Figure 7 The example shown can be executed by an edge server. The edge server can be communicatively connected to a radar device and a camera device respectively. The radar device is used to monitor the storage goods within the field of view in real time, and the camera device is used to take real-time pictures of the storage goods. In addition, the edge server can also be communicatively connected to a cloud server. The cloud server can be configured with various services, such as a data storage server, a visual detection algorithm service, etc. Thus, the edge server can call the services configured in the cloud server to perform corresponding operations.

[0101] As Figure 7 shown, at 710, the point cloud data within the field of view collected by the radar device is obtained.

[0102] At 720, compare the acquired adjacent two-frame point cloud data, and the frame interval time of the adjacent two-frame point cloud data is a specified duration.

[0103] At 730, it is possible to determine whether there is cargo movement based on the comparison result at 720. If there is cargo movement, perform the operation at 740; if there is no cargo movement, return to perform the operation at 710, that is, continue to acquire the point cloud data collected by the radar device to monitor whether there is cargo movement.

[0104] At 740, acquire two frames of images corresponding to the two-frame point cloud data, and the acquisition times of the two-frame cloud data and the corresponding two frames of images are the same.

[0105] At 750, call the visual detection algorithm configured by the cloud server to determine the moving cargo in the warehoused goods based on the two frames of images.

[0106] In one example, comparing the acquired adjacent two-frame point cloud data to determine whether there is cargo movement includes: substituting the acquired adjacent two-frame point cloud data into the pre-divided first field-of-view division space respectively to obtain the tree structure indexes corresponding to the two-frame point cloud data respectively, where the first field-of-view division space is obtained by dividing the space within the field of view using the octree algorithm based on the point cloud data generated by the radar device, and the tree structures of the two obtained tree structure indexes are the same; and determining whether there is cargo movement based on the tree structure indexes corresponding to the two-frame point cloud data.

[0107] In one example, comparing the acquired adjacent two-frame point cloud data to determine whether there is cargo movement includes: using the octree algorithm to divide the space within the field of view based on one of the two-frame point cloud data to generate a second field-of-view division space and obtain the tree structure index corresponding to this frame of point cloud data; substituting the other frame of point cloud data in the two-frame point cloud data into the second field-of-view division space to obtain the tree structure index corresponding to the other frame of point cloud data, where the tree structures of the two obtained tree structure indexes are the same; and determining whether there is cargo movement based on the tree structure indexes corresponding to the two-frame point cloud data.

[0108] In one example, determining whether there is cargo movement based on the tree structure indexes corresponding to the two-frame point cloud data includes: respectively comparing the leaf nodes at the end of each branch in the tree structure indexes corresponding to the two-frame point cloud data to count the number of different leaf nodes in the two tree structure indexes; and when the number of different leaf nodes is greater than the second quantity threshold, determining that there is cargo movement.

[0109] In one example, the leaf nodes in the tree - like structure indexes corresponding to two frames of point cloud data are respectively compared to determine the number of different leaf nodes in the tree - like structure indexes, including: for two leaf nodes in the tree - like structure indexes corresponding to two frames of point cloud data that represent the same three - dimensional space, comparing the point cloud data included in the three - dimensional spaces represented by the two leaf nodes; when the number of different point cloud data is greater than the third quantity threshold, determining that the two leaf nodes are different; and when the number of different point cloud data is not greater than the third quantity threshold, determining that the two leaf nodes are the same.

[0110] In one example, before comparing the adjacent two frames of point cloud data obtained, the cargo movement detection method may further include: performing filtering processing on the adjacent two frames of point cloud data obtained.

[0111] In one example, the cargo movement detection method may further include: uploading the adjacent two frames of cloud data, two frames of images, and information about the moving cargo to the blockchain.

[0112] Figure 8 FIG. shows a block diagram of an example of a device for detecting cargo movement according to an embodiment of the present specification (hereinafter referred to as the cargo movement detection device 800).

[0113] The cargo movement detection device 800 can be applied to an edge server. The edge server is respectively communicatively connected to a radar device and a camera device, and is also communicatively connected to a cloud server. The radar device is used to monitor the storage cargo within the field of view in real - time, and the camera device is used to capture the storage cargo in real - time.

[0114] The cargo movement detection device 800 includes: a point cloud data acquisition unit 810, a point cloud data comparison unit 820, an image acquisition unit 830, and an algorithm call unit 840.

[0115] The point cloud data acquisition unit 810 can be configured to acquire the point cloud data within the field of view collected by the radar device.

[0116] The point cloud data comparison unit 820 is configured to compare the adjacent two frames of point cloud data obtained to determine whether there is cargo movement; when it is determined that there is cargo movement, it can trigger the image acquisition unit 830. The frame interval time of the adjacent two frames of point cloud data is a specified duration.

[0117] The image acquisition unit 830 can be configured to acquire two frames of images corresponding to the two frames of point cloud data, and the acquisition times of the two frames of cloud data and the corresponding two frames of images are the same.

[0118] The algorithm call unit 840 can be configured to call the visual detection algorithm configured by the cloud server to determine the moving cargo in the storage cargo according to the two frames of images.

[0119] In one example, the point cloud data acquisition unit 810 may also be configured to: respectively substitute two adjacent frames of acquired point cloud data into a pre-divided first field of view division space to obtain tree structure indexes corresponding to the two frames of point cloud data, where the first field of view division space is obtained by dividing the space within the field of view according to the point cloud data generated by the radar device using the octree algorithm, and the tree structures of the two obtained tree structure indexes are the same; and determine whether there is cargo movement according to the tree structure indexes corresponding to the two frames of point cloud data.

[0120] In one example, the point cloud data acquisition unit 810 may also be configured to: use the octree algorithm to divide the space within the field of view according to one of the two frames of point cloud data to generate a second field of view division space and obtain the tree structure index corresponding to this frame of point cloud data; substitute the other frame of point cloud data among the two frames of point cloud data into the second field of view division space to obtain the tree structure index corresponding to the other frame of point cloud data, where the tree structures of the two obtained tree structure indexes are the same; and determine whether there is cargo movement according to the tree structure indexes corresponding to the two frames of point cloud data.

[0121] In one example, the point cloud data acquisition unit 810 may also be configured to: respectively compare the leaf nodes at the end of each branch in the tree structure indexes corresponding to the two frames of point cloud data to count the number of different leaf nodes in the two tree structure indexes; and when the number of different leaf nodes is greater than the second quantity threshold, determine that there is cargo movement.

[0122] In one example, the point cloud data acquisition unit 810 may also be configured to: for two leaf nodes in the tree structure indexes corresponding to the two frames of point cloud data that represent the same three-dimensional space, compare the point cloud data included in the three-dimensional spaces represented by the two leaf nodes; when the number of different point cloud data is greater than the third quantity threshold, determine that the two leaf nodes are different; and when the number of different point cloud data is not greater than the third quantity threshold, determine that the two leaf nodes are the same.

[0123] In one example, the cargo movement detection device 800 may also include a filtering unit, and the filtering unit may be configured to: perform filtering processing on the two adjacent frames of acquired point cloud data.

[0124] In one example, the cargo movement detection device 800 may also include an uploading unit, and the uploading unit may be configured to: upload two adjacent frames of cloud data, two frames of images, and information about the moving cargo.

[0125] The above refers to Figures 1 to 8 and describes the embodiments of the method and device for detecting cargo movement according to the embodiments of this specification.

[0126] The device for detecting the movement of goods in the embodiments of this specification can be implemented in hardware, or can be implemented by software or a combination of hardware and software. Taking software implementation as an example, as a logically meaningful device, it is formed by the processor of the device where it is located reading the corresponding computer program instructions in the memory into the memory for operation. In the embodiments of this specification, the device for detecting the movement of goods can be implemented by using an electronic device, for example.

[0127] Figure 9 The block diagram of an example of an electronic device 900 for implementing the method for detecting the movement of goods according to the embodiments of this specification is shown.

[0128] As Figure 9 shown, the electronic device 900 may include at least one processor 910, a memory (for example, a non-volatile memory) 920, a memory 930, and a communication interface 940, and at least one processor 910, the memory 920, the memory 930, and the communication interface 940 are connected together via a bus 950. At least one processor 910 executes at least one computer-readable instruction stored or encoded in the memory (that is, the above elements implemented in software form).

[0129] In one embodiment, computer-executable instructions are stored in the memory, which when executed cause at least one processor 910 to: obtain point cloud data within the field of view collected by a radar device; compare two adjacent frames of the obtained point cloud data to determine whether there is movement of goods; when it is determined that there is movement of goods, obtain two frames of images corresponding to the two frames of point cloud data, and the acquisition times of the two frames of cloud data and the corresponding two frames of images are the same; and call the visual detection algorithm configured by the cloud server to determine the moving goods in the warehoused goods according to the two frames of images.

[0130] It should be understood that the computer-executable instructions stored in the memory, when executed, cause at least one processor 910 to perform the various operations and functions described above in the various embodiments of this specification in combination with Figures 1 - 8 the description.

[0131] According to one embodiment, a program product such as a machine-readable medium is provided. The machine-readable medium may have instructions (that is, the above elements implemented in software form), which when executed by the machine, cause the machine to perform the various operations and functions described above in the various embodiments of this specification in combination with Figures 1 - 8 the description.

[0132] Specifically, a system or device equipped with a readable storage medium can be provided, on which software program codes for implementing the functions of any one of the above embodiments are stored, and the computer or processor of the system or device is caused to read and execute the instructions stored in the readable storage medium.

[0133] In this case, the program code read from the readable medium itself can implement the functions of any one of the above embodiments, so the machine-readable code and the readable storage medium storing the machine-readable code constitute a part of the present invention.

[0134] The computer program codes required for the operations of the various parts of this specification can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB, NET, and Python, conventional procedural programming languages such as C, Visual Basic 2003, Perl, COBOL 2002, PHP, and ABAP, dynamic programming languages such as Python, Ruby, and Groovy, or other programming languages. The program code can run on the user's computer, or run on the user's computer as a stand-alone software package, or part of it runs on the user's computer and another part runs on a remote computer, or all of it runs on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer in any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet), or in a cloud computing environment, or used as a service, such as software as a service (SaaS).

[0135] Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tapes, non-volatile memory cards, and ROMs. Optionally, the program code can be downloaded from a server computer or the cloud via a communication network.

[0136] The above has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0137] Not all steps and units in the above-mentioned processes and system structure diagrams are necessary, and some steps or units can be ignored according to actual needs. The execution order of each step is not fixed and can be determined as required. The device structures described in the above-mentioned embodiments can be physical structures or logical structures, that is, some units may be implemented by the same physical entity, or some units may be implemented by multiple physical entities separately, or some components in multiple independent devices may be jointly implemented.

[0138] The term "exemplary" used throughout this specification means "serving as an example, instance, or illustration", and does not mean "preferred" or "advantageous" compared to other embodiments. For the purpose of providing an understanding of the described technology, the detailed description includes specific details. However, these technologies can be implemented without these specific details. In some instances, well-known structures and devices are shown in block diagram form to avoid obscuring the concepts of the described embodiments.

[0139] The optional implementation manners of the embodiments of this specification have been described in detail above in conjunction with the accompanying drawings. However, the embodiments of this specification are not limited to the specific details in the above-mentioned implementation manners. Within the scope of the technical concept of the embodiments of this specification, various simple modifications can be made to the technical solutions of the embodiments of this specification, and these simple modifications all fall within the protection scope of the embodiments of this specification.

[0140] The above description of the content of this specification is provided to enable any ordinary person skilled in the art to implement or use the content of this specification. For ordinary persons skilled in the art, various modifications to the content of this specification are obvious, and the general principles defined herein can also be applied to other variations without departing from the protection scope of the content of this specification. Therefore, the content of this specification is not limited to the examples and designs described herein, but is consistent with the broadest scope that conforms to the principles and novel features disclosed herein.

Claims

1. A method for detecting the movement of goods, comprising: acquiring point cloud data within the field of view collected by a radar device, where the radar device is used for real-time monitoring of storage goods within the field of view; comparing two adjacent frames of the acquired point cloud data to determine whether there is goods movement, where the frame interval time of the two adjacent frames of point cloud data is a specified duration; when it is determined that there is goods movement, acquiring two frames of images corresponding to the two frames of point cloud data, where the images are collected by a camera device through real-time photographing of the storage goods, and the acquisition times of the two frames of point cloud data and the corresponding two frames of images are the same; and determining the moving goods in the storage goods according to the two frames of images, where comparing the two adjacent frames of the acquired point cloud data to determine whether there is goods movement includes: using an octree algorithm to perform spatial partitioning on the space within the field of view according to one of the two frames of point cloud data to generate a second field-of-view partitioned space and obtain a tree-like structure index corresponding to this frame of point cloud data; substituting the other frame of point cloud data of the two frames of point cloud data into the second field-of-view partitioned space to obtain a tree-like structure index corresponding to the other frame of point cloud data, where the tree-like structures of the two obtained tree-like structure indexes are the same; and determining whether there is goods movement according to the tree-like structure indexes corresponding to the two frames of point cloud data.

2. The method according to claim 1, wherein, determining whether there is goods movement according to the tree-like structure indexes corresponding to the two frames of point cloud data includes: respectively comparing the leaf nodes at the end of each branch in the tree-like structure indexes corresponding to the two frames of point cloud data to count the number of different leaf nodes in the two tree-like structure indexes; and when the number of different leaf nodes is greater than a second quantity threshold, determining that there is goods movement.

3. The method according to claim 2, wherein, respectively comparing the leaf nodes in the tree-like structure indexes corresponding to the two frames of point cloud data to determine the number of different leaf nodes in the tree-like structure indexes includes: for two leaf nodes in the tree-like structure indexes corresponding to the two frames of point cloud data that represent the same three-dimensional space, comparing the point cloud data included in the three-dimensional space represented by the two leaf nodes; when the number of different point cloud data is greater than a third quantity threshold, determining that the two leaf nodes are different; and when the number of different point cloud data is not greater than the third quantity threshold, determining that the two leaf nodes are the same.

4. The method according to claim 1, wherein, the spatial partitioning performed using the octree algorithm according to the point cloud data generated by the radar device is executed in the following manner: for each three-dimensional space to be partitioned in the space within the field of view, determining whether the number of point clouds included in the three-dimensional space to be partitioned is greater than a first quantity threshold; if it is greater, partitioning the three-dimensional space to be partitioned into 8 quadrants to obtain 8 three-dimensional spaces to be partitioned; and if it is not greater, no longer partitioning the three-dimensional space to be partitioned.

5. The method according to claim 1, wherein, Before comparing the acquired adjacent two-frame point cloud data, the method further includes: Performing filtering processing on the acquired adjacent two-frame point cloud data.

6. The method according to claim 1, wherein, Determining the moving goods in the storage goods according to the two frames of images includes: Processing the two frames of images by using a visual detection algorithm to determine the moving goods in the storage goods.

7. The method according to claim 1, further including: Uploading the adjacent two-frame point cloud data, the two frames of images, and the information of the moving goods to the blockchain.

8. The method according to claim 1, further including: Issuing a warning when the time point of goods movement belongs to a specified abnormal time period.

9. A method for detecting goods movement, which is executed by an edge server. The edge server is respectively communicatively connected to a radar device and a camera device, and is also communicatively connected to a cloud server. The radar device is used for real-time monitoring of storage goods within a field of view, and the camera device is used for real-time photographing of the storage goods. The method includes: Obtaining the point cloud data within the field of view collected by the radar device; Comparing the acquired adjacent two-frame point cloud data to determine whether there is goods movement. The frame interval time of the adjacent two-frame point cloud data is a specified duration; When it is determined that there is goods movement, obtaining two frames of images corresponding to the two-frame point cloud data. The acquisition times of the two-frame point cloud data and the corresponding two frames of images are the same; and Invoking the visual detection algorithm configured by the cloud server to determine the moving goods in the storage goods according to the two frames of images, wherein, comparing the acquired adjacent two-frame point cloud data to determine whether there is goods movement includes: Using an octree algorithm to perform spatial partitioning on the space within the field of view according to one of the two-frame point cloud data to generate a second field-of-view partitioning space and obtain a tree-like structure index corresponding to this frame of point cloud data; Substituting the other frame of point cloud data in the two-frame point cloud data into the second field-of-view partitioning space to obtain a tree-like structure index corresponding to the other frame of point cloud data, wherein the tree-like structures of the two obtained tree-like structure indexes are the same; and Determining whether there is goods movement according to the tree-like structure indexes corresponding to the two-frame point cloud data.

10. The method according to claim 9, wherein, Determining whether there is goods movement according to the tree-like structure indexes corresponding to the two-frame point cloud data includes: Respectively comparing the leaf nodes at the end of each branch in the tree-like structure indexes corresponding to the two-frame point cloud data to count the number of different leaf nodes in the two tree-like structure indexes; and When the number of different leaf nodes is greater than a second quantity threshold, determining that there is goods movement.

11. The method according to claim 10, wherein, Respectively comparing the leaf nodes in the tree-like structure indexes corresponding to the two-frame point cloud data to determine the number of different leaf nodes in the tree-like structure indexes includes: For two leaf nodes used to represent the same three-dimensional space in the tree - like structure index corresponding to the two frames of point cloud data, compare the point cloud data included in the three - dimensional spaces represented by the two leaf nodes; When the number of different point cloud data is greater than a third quantity threshold, determine that the two leaf nodes are different; and When the number of different point cloud data is not greater than the third quantity threshold, determine that the two leaf nodes are the same.

12. A device for detecting the movement of goods, which is applied to an edge server. The edge server is respectively communicatively connected to a radar device and a camera device, and is also communicatively connected to a cloud server. The radar device is used to monitor the storage goods within the field of view in real time, and the camera device is used to capture the storage goods in real time. The device comprises: A point cloud data acquisition unit configured to acquire the point cloud data within the field of view collected by the radar device; A point cloud data comparison unit configured to compare two adjacent frames of point cloud data obtained to determine whether there is goods movement. When it is determined that there is goods movement, trigger the image acquisition unit. The frame interval time of the two adjacent frames of point cloud data is a specified duration; The image acquisition unit configured to acquire two frames of images corresponding to the two frames of point cloud data, and the acquisition times of the two frames of point cloud data and the corresponding two frames of images are the same; and An algorithm call unit configured to call the visual detection algorithm configured by the cloud server to determine the moving goods in the storage goods according to the two frames of images. Wherein, the point cloud data comparison unit is configured to: Use the octree algorithm to divide the space within the field of view according to one of the two frames of point cloud data to generate a second field - of - view division space and obtain the tree - like structure index corresponding to this frame of point cloud data; Substitute the other frame of point cloud data in the two frames of point cloud data into the second field - of - view division space to obtain the tree - like structure index corresponding to the other frame of point cloud data, wherein the tree - like structures of the two obtained tree - like structure indexes are the same; and Determine whether there is goods movement according to the tree - like structure indexes corresponding to the two frames of point cloud data.

13. An electronic device comprises: At least one processor, a memory coupled to the at least one processor, and a computer program stored on the memory. The at least one processor executes the computer program to implement the method according to any one of claims 9 - 11.

14. A computer - readable storage medium storing a computer program, where the computer program, when executed by a processor, implements the method according to any one of claims 9 - 11.

15. A computer program product comprising a computer program, where the computer program, when executed by a processor, implements the method according to any one of claims 9 - 11.

Images