Data acquisition method and data acquisition apparatus based on 3D map

Abstract

The application provides a 3D map-based data collection method and a data collection device. The method comprises: acquiring description data of a target region of an open-pit mine, and determining a plurality of undetermined member information corresponding to the description data according to the description data, the description data being data for describing information of the target region from one dimension, and one undetermined member information being used for representing one characteristic information of the description data; determining a target feature vector corresponding to the undetermined member according to a correlation between any two undetermined members in the plurality of undetermined member information, the target feature vector being used for representing a vectorization result of the characteristic information; performing clustering processing on the target feature vector to obtain at least one clustering center vector, and determining that the undetermined member information corresponding to the target feature vector is target member information; and acquiring description data corresponding to the target member information to obtain target data. The problem of low precision of the prior art data collection method is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a data collection method and device based on a 3D map. BACKGROUND

[0002] An open-pit mine 3D map is a map that uses three-dimensional technology to display an open-pit mine site. It presents information such as the terrain, mining area, and excavation equipment of the mine through visualization and simulation technology. Such a map can provide a more intuitive and vivid way to understand and analyze the structure, scale, and mining conditions of the mine. Since the open-pit mine site changes over time during development, the open-pit mine 3D map needs to be updated so that mining personnel can accurately grasp the current information of the open-pit mine. Therefore, the current regional information of the open-pit mine site needs to be obtained to update the constructed open-pit mine 3D map based on the regional information. To reduce the amount of data processing during the updating process, the collected information of multiple dimensions is filtered to extract target information from the initial information. However, in the prior art, the data collection process has relatively low precision. SUMMARY

[0003] The main purpose of the present application is to provide a data collection method and device based on a 3D map to at least solve the problem of low precision of the data collection method in the prior art.

[0004] According to an aspect of the present application, a data collection method based on a 3D map is provided, including: obtaining description data of a target region of an open-pit mine, and determining a plurality of pending member information corresponding to the description data according to the description data, wherein the description data is data describing information of the target region from one dimension, and one pending member information is used to represent a feature information of the description data; determining a target feature vector corresponding to the pending members according to the correlation between any two of the pending members in the plurality of pending member information, wherein the target feature vector is used to represent the vectorization result of the feature information; performing clustering processing on the target feature vector to obtain at least one cluster center vector, and determining the pending member information corresponding to the target feature vector as target member information; obtaining the description data corresponding to the target member information to obtain target data.

[0005] Optionally, the description data is image data or text data, in the case of the description data being the image data, according to the description data, the plurality of pending member information corresponding to the description data is determined, comprising: obtaining at least one description image in the description data, wherein the description image is obtained by image information collection on the target area; obtaining first feature information corresponding to the description image, and determining the pending member information corresponding to the description data according to the first feature information; in the case of the description data being the text data, according to the description data, the plurality of pending member information corresponding to the description data is determined, comprising: obtaining at least one text segment in the description data, wherein the text segment is obtained by text recording on the mining process in the target area; marking the text segment as second feature information, and determining the pending member information corresponding to the description data according to the second feature information.

[0006] Optionally, according to the correlation between any two of the plurality of pending members, the target feature vector corresponding to the pending member is determined, comprising: generating a member relationship network according to the correlation between any two of the plurality of pending members; determining the target feature vector corresponding to the pending member according to the member relationship network and a cycle optimization rule.

[0007] Optionally, according to the member relationship network and the cycle optimization rule, the target feature vector corresponding to the pending member is determined, comprising: an obtaining step of obtaining a first initial feature vector of a neighboring member of the pending member, wherein the neighboring member is linked with the pending member, and the distribution order of the neighboring member is after the distribution order of the pending member; a calculation step of calculating an association parameter between the pending member and the neighboring member according to the first initial feature vector to obtain a first association parameter; a processing step of optimizing the first initial feature vector according to the first association parameter to obtain an intermediate feature vector; a repeating step of repeating the obtaining step, the calculation step and the processing step at least once, and in the repeating process, updating the first initial feature vector in the obtaining step to the intermediate feature vector in the processing step obtained in the last repeating process, and determining that the intermediate feature vector under the condition of reaching a preset condition is the target feature vector, wherein the preset condition is reaching a preset number of times.

[0008] Optionally, the processing step comprises: obtaining a second initial feature vector of another pending member of the pending member, wherein the other pending member is located in the same member relationship network as the pending member and is different from the pending member; calculating an association parameter between the pending member and the adjacent member according to the initial feature vector to obtain a second association parameter; and performing weighting processing on the second initial feature vector according to the second association parameter to obtain the intermediate feature vector.

[0009] Optionally, the member relationship network is generated according to the correlation between any two of the plurality of pending members, comprising: obtaining distribution ordering information between the pending members, wherein the distribution ordering information is used to represent the distribution ordering between the pending members, the pending members corresponding to the same description data have the same distribution ordering, and the pending members corresponding to different description data have different member distribution orderings; determining a local area position corresponding to the pending member according to feature information corresponding to the pending member, wherein the area granularity corresponding to the feature information is greater than or equal to the area granularity of the local area position; performing deduplication processing on the local area position to obtain at least one target local position, and marking the at least one target local position to obtain a link relationship member; determining member link information between the pending members according to the link relationship member, wherein the member link information is used to represent the link relationship between the pending members, the pending members having different distribution orderings are linked through at least one link relationship member, for any one of the pending members linked by any one of the link relationship members, the local area position corresponding to the feature information corresponding to the pending member includes the target local position corresponding to the link relationship member, and the link relationship members have the same distribution ordering in the distribution ordering information; and forming a corresponding member relationship network according to each of the pending members, the distribution ordering information and the member link information.

[0010] Optionally, after forming the corresponding member relationship network according to each of the pending members, the distribution ordering information and the member link information, and before determining the target feature vector corresponding to the pending member according to the member relationship network and the cycle optimization rule, the method further comprises: updating the distribution ordering of the pending members according to the distribution ordering information.

[0011] Optionally, the calculation step further comprises: obtaining a statistical distribution parameter of the target local position of the link relationship member; and calculating a link edge association parameter to obtain the first association parameter according to the statistical distribution parameter, wherein the link edge association parameter is used to represent the link relationship between the link relationship member and the pending member.

[0012] Optionally, the description data corresponding to the target member information is acquired to obtain target data, including: acquiring the description data corresponding to a plurality of the target member information to obtain a plurality of preliminary data; and performing combination processing on the plurality of preliminary data to obtain the target data.

[0013] According to another aspect of the present application, a data acquisition device based on a 3D map is provided, including: a first acquisition unit configured to acquire description data of a target region of an open-pit mine, and determine a plurality of undetermined member information corresponding to the description data according to the description data, wherein the description data is data describing information of the target region from one dimension, and one of the undetermined member information is used to represent a kind of characteristic information of the description data; a first determination unit configured to determine a target feature vector corresponding to the undetermined member according to a correlation between any two of the undetermined members in the plurality of undetermined member information, wherein the target feature vector is used to represent a vectorization result of the characteristic information; a second determination unit configured to perform clustering processing on the target feature vector to obtain at least one cluster center vector, and determine the undetermined member information corresponding to the target feature vector as target member information; and a second acquisition unit configured to acquire the description data corresponding to the target member information to obtain target data.

[0014] By applying the technical solution of the present application, first, the description data of the target region of the open-pit mine is acquired, and a plurality of undetermined member information corresponding to the description data is determined according to the description data, wherein the description data is data describing information of the target region from one dimension, and one of the undetermined member information is used to represent a kind of characteristic information of the description data; then, a target feature vector corresponding to the undetermined member is determined according to a correlation between any two of the undetermined members in the plurality of undetermined member information, wherein the target feature vector is used to represent a vectorization result of the characteristic information; then, clustering processing is performed on the target feature vector to obtain at least one cluster center vector, and the undetermined member information corresponding to the target feature vector is determined as target member information; finally, the description data corresponding to the target member information is acquired to obtain target data. In the above solution, since the target open-pit mine region data is selected from a plurality of open-pit mine description data, not only one kind of open-pit mine description data is refined into at least one corresponding undetermined relationship network member, i.e., one kind of characteristic information, but also the granularity of the analysis object is smaller, and the data acquisition accuracy is further improved, so that the data acquisition accuracy for constructing the 3D map of the open-pit mine can be effectively improved. The problem of low accuracy of the data acquisition method based on the 3D map in the prior art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application and are incorporated in and constitute a part of this application. The illustrations, together with the description, serve to explain the application, but are not intended to limit the application in any way.

[0016] Figure 1 A hardware structure block diagram of a mobile terminal performing a 3D map-based data collection method is shown according to an embodiment of the present application;

[0017] Figure 2 A flowchart of a 3D map-based data collection method is shown according to an embodiment of the present application;

[0018] Figure 3 A schematic diagram of a member relationship network is shown according to an embodiment of the present application;

[0019] Figure 4 A structure block diagram of a 3D map-based data collection apparatus is shown according to an embodiment of the present application.

[0020] Among the above-mentioned drawings, the following reference signs are included:

[0021] 102, processor; 104, memory; 106, transmission device; 108, input / output device. DETAILED DESCRIPTION

[0022] It should be noted that the embodiments and features of the embodiments in the present application can be combined with each other without conflict. The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0023] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] As described in the background section, existing data acquisition methods have low accuracy. To address this issue, embodiments of this application provide a data acquisition method and device based on 3D maps.

[0026] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0027] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a data acquisition method based on a 3D map, according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0028] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the 3D map-based data acquisition method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0029] This embodiment provides a 3D map-based data acquisition method that runs on a mobile terminal, computer terminal, or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] Figure 2 This is a flowchart of a 3D map-based data acquisition method according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0031] Step S201: Obtain description data of the target area of ​​the open-pit mine, and determine multiple undetermined member information corresponding to the description data based on the description data. The description data is data that describes the information of the target area from one dimension, and one of the undetermined member information is used to characterize a feature information of the description data.

[0032] Specifically, at least one pending member corresponding to each open-pit mine description data can be determined according to the determined plurality of open-pit mine description data. Each pending member is used to reflect a feature information included in the corresponding open-pit mine description data, that is, a key data in the corresponding open-pit mine description data. One open-pit mine description data can correspond to one or more key data, that is, one or more pending members. The plurality of open-pit mine description data is used to respectively characterize the regional information of the target open-pit mine area from multiple dimensions. That is, one open-pit mine description data is used to respectively characterize the regional information of the target open-pit mine area from one dimension. The dimension can refer to the way of forming the open-pit mine description data, or the dimension can also refer to the data form of the open-pit mine description data, such as image data, text data, etc. For example, one frame of image is obtained by image acquisition of one position in the target open-pit mine area. For another example, the key data of one open-pit mine description data is formed by recording the mining process of a sub-region of the target open-pit mine area.

[0033] In step S202, a target feature vector corresponding to each pending member is determined according to the correlation between any two of the plurality of pending member information. The target feature vector is used to represent the vectorization result of the feature information.

[0034] Specifically, the feature information reflected by each pending member can be vectorized. Then, the correlation between the pending members can be mined based on the vectorization result to optimize the initial vector and obtain the target feature vector corresponding to each pending member, so that the representation ability of the target feature vector is better. For example, the pending member in the image dimension learns the feature information of the pending member in the text dimension.

[0035] In step S203, the target feature vector is clustered to obtain at least one cluster center vector, and the pending member information corresponding to the target feature vector is determined as target member information.

[0036] Specifically, whether the corresponding target feature vector is representative or whether it meets other conditions can be used to determine whether the pending relationship network member belongs to the target member. Through the above method, at least one target member, such as image A or text segment 1, can be determined.

[0037] In step S204, the description data corresponding to the target member information is obtained to obtain target data.

[0038] Specifically, the target data is used to update the 3D map of the open-pit mine constructed for the target open-pit mine area. The specific map updating method can adopt the prior art. The application focuses on how to collect and screen data, and therefore, is not specifically limited in the application. In some embodiments, the target data can also be used as a basis for map updating. For example, similarity calculation can be performed based on the current target open-pit mine area data and the historical target open-pit mine area data. The similarity calculation can be any data similarity calculation, such as text similarity and image similarity. The similarities of different dimensions of data are fused, such as mean value. When the obtained data similarity is less than a preset similarity, it can be determined that the mining area has changed greatly, and map updating is required. When the obtained data similarity is greater than or equal to the preset similarity, it can be determined that the mining area has changed slightly, and map updating is not required.

[0039] Suppose there are three types of description data: image sequence A, image sequence B and text data 1. The determined pending members are: the pending member corresponding to image sequence A, pending member 1 and pending member 2; the pending member corresponding to image sequence B: pending member 3 and pending member 4; the pending member corresponding to text data 1: pending member 5 and pending member 6. The corresponding member correlation formed by the determined pending members can be: there is a correlation between pending member 1 and pending member 3; there is a correlation between pending member 2 and pending member 4; there is a correlation between pending member 3 and pending member 6; there is a correlation between pending member 4 and pending member 5. The target feature vector corresponding to pending member 1 is feature vector 1, the target feature vector corresponding to pending member 2 is feature vector 2, the target feature vector corresponding to pending member 3 is feature vector 3, the target feature vector corresponding to pending member 4 is feature vector 4, and the target feature vector corresponding to pending member 5 is feature vector 5. The target feature vector corresponding to pending member 6 is feature vector 6. Suppose the pending members corresponding to feature vector 1 and feature vector 2 are determined as target members, and the target data can be determined based on the pending members corresponding to the two feature vectors.

[0040] By the embodiment, first, description data of a target region of an open-pit mine is acquired, and a plurality of pending member information corresponding to the description data is determined according to the description data, the description data being data describing information of the target region from one dimension, and one pending member information being used to represent one characteristic information of the description data; then, a target feature vector corresponding to the pending members is determined according to a correlation between any two pending members in the plurality of pending member information, the target feature vector being used to represent a vectorization result of the characteristic information; then, the target feature vector is subjected to clustering processing to obtain at least one clustering center vector, and the pending member information corresponding to the target feature vector is determined as target member information; finally, description data corresponding to the target member information is acquired to obtain target data. In the above scheme, in the process of screening the target open-pit mine region data from a plurality of open-pit mine description data, one open-pit mine description data is refined into at least one corresponding pending member, i.e., one characteristic information, so that the granularity of the analysis object is smaller, and the data acquisition accuracy is further improved, thereby effectively improving the data acquisition accuracy for the construction of the open-pit mine 3D map. The problem of low accuracy of the prior art data acquisition method based on the 3D map is solved.

[0041] In the specific implementation process, the description data is image data or text data, in the case where the description data is the image data, the step S201 can be implemented through the following steps: a step S2011 of acquiring at least one description image in the description data, wherein the description image is obtained by image information acquisition on the target region; and a step S2012 of acquiring first characteristic information corresponding to the description image, and determining the pending member information corresponding to the description data according to the first characteristic information; in the case where the description data is the text data, the step S201 can also be implemented through the following steps: a step S2013 of acquiring at least one text segment in the description data, wherein the text segment is obtained by text recording on the mining process in the target region; and a step S2014 of marking the text segment as second characteristic information, and determining the pending member information corresponding to the description data according to the second characteristic information. The method can further quickly determine the plurality of pending member information corresponding to the description data.

[0042] Specifically, in the case of the above description data being the above image data, in the above one image sequence, each open-pit mine description image such as an image of a mining pit or an image of an angle in a smaller granularity mining pit can be extracted. Each open-pit mine description image is taken as one kind of feature information, and 3D image acquisition (i.e., a depth image) can be exemplarily performed by a flight equipment. In the case of the above description data being the above text data, in the above one description data, at least one data segment such as a data of a region is determined, and the above data segment is formed by recording the mining process in the above target region to obtain a recorded text of each region.

[0043] To further improve the reliability of data acquisition, the above step S202 can be implemented by the following steps: step S2021, generating a member relationship network according to the correlation between any two of the above pending members; and step S2022, determining the target feature vector corresponding to the above pending member according to the above member relationship network and a cyclic optimization rule. This method can cyclically update the correlation between the pending members by generating a member relationship network and combining a cyclic optimization rule, and then learn the information of another dimension of the pending member, which can enhance the representation ability of the target feature vector, so as to effectively improve the reliability of data acquisition for open-pit mine 3D map construction.

[0044] Specifically, the above member relationship network can be used to reflect the correlation between the determined pending members.

[0045] The above step S2022 can be implemented in other ways, for example: step S20221, an obtaining step, obtaining a first initial feature vector of a neighboring member of the above pending member, wherein the above neighboring member is linked with the above pending member, and the distribution order of the above neighboring member is after the distribution order of the above pending member; step S20222, a calculation step, calculating an association parameter between the above pending member and the above neighboring member according to the above first initial feature vector to obtain a first association parameter; step S20223, a processing step, optimizing the above first initial feature vector according to the above first association parameter to obtain an intermediate feature vector; and step S20224, a repeating step, repeating the above obtaining step, the above calculation step, and the above processing step at least once, and in the repeating process, updating the above first initial feature vector in the above obtaining step to the above intermediate feature vector in the above processing step obtained in the last repeating process, and determining the above intermediate feature vector under the condition that a preset condition is reached as the above target feature vector, wherein the above preset condition is reaching a preset number of times. This method can further compensate for the limitations of information of different dimensions in representing information.

[0046] Specifically, the initial feature vector of the undetermined network member can be obtained by performing convolution operation on the feature information corresponding to the undetermined member, wherein the feature information of the image dimension and the feature information of the text dimension can be realized by different neural networks, such as different convolution neural networks, and the specific implementation is not limited, and the existing processing process for feature extraction or feature mining of images and texts can be adopted.

[0047] For example, for the first optimization: the initial open pit feature vector of the undetermined member 1 is taken as the intermediate open pit feature vector after the first optimization; the initial open pit feature vector of the undetermined member 2 is taken as the intermediate open pit feature vector after the first optimization; the initial open pit feature vector of the undetermined member 3 is taken as the intermediate open pit feature vector after the first optimization; for the undetermined member 4, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1, the undetermined member 2 and the undetermined member 3), the initial open pit feature vector of the undetermined member 4 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 4; for the undetermined member 5, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1 and the undetermined member 3), the initial open pit feature vector of the undetermined member 5 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 5; for the undetermined member 6, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1, the undetermined member 2 and the undetermined member 3), the initial open pit feature vector of the undetermined member 6 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 6.

[0048] For the second optimization: the intermediate open-pit mine feature vector of the first optimization of the pending member 4 is taken as the intermediate open-pit mine feature vector after the second optimization; the intermediate open-pit mine feature vector of the first optimization of the pending member 5 is taken as the intermediate open-pit mine feature vector after the second optimization; the intermediate open-pit mine feature vector of the first optimization of the pending member 6 is taken as the intermediate open-pit mine feature vector after the second optimization; for the pending member 1, based on the intermediate open-pit mine feature vectors of the related pending members (i.e. the intermediate open-pit mine feature vectors after the first optimization of the pending member 4, the pending member 5 and the pending member 6), the intermediate open-pit mine feature vector of the first optimization of the pending member 1 is optimized to form the intermediate open-pit mine feature vector after the second optimization of the pending member 1; for the pending member 2, based on the intermediate open-pit mine feature vectors of the related pending members (i.e. the intermediate open-pit mine feature vectors after the first optimization of the pending member 4 and the pending member 6), the intermediate open-pit mine feature vector of the first optimization of the pending member 2 is optimized to form the intermediate open-pit mine feature vector after the second optimization of the pending member 2; for the pending member 3, based on the intermediate open-pit mine feature vectors of the related pending members (i.e. the intermediate open-pit mine feature vectors after the first optimization of the pending member 4 and the pending member 6), the intermediate open-pit mine feature vector of the first optimization of the pending member 3 is optimized to form the intermediate open-pit mine feature vector after the second optimization of the pending member 3. Through the continuous circulation of the above optimization steps, the optimization of the two mutually inverse sequential ordering relationships from top to bottom and from bottom to top can be realized, and through the successive circulation, the feature information between the pending members in the above member relationship network can be fully interacted, and the number of specific circulation is not limited.

[0049] The step S20223 can also be implemented in other ways, for example: step S202231, obtaining a second initial feature vector of another pending member of the pending member, wherein the other pending member is located in the same member relationship network as the pending member, and the other pending member is different from the pending member; step S202232, calculating the correlation parameter between the pending member and the adjacent member according to the initial feature vector to obtain a second correlation parameter; step S202233, weighting the second initial feature vector according to the second correlation parameter to obtain the intermediate feature vector. This method can further quickly calculate the intermediate feature vector.

[0050] Specifically, for each pending member related to the above-mentioned related next pending member, including the above-mentioned one pending member, at least one other member related to the above-mentioned related next pending member, the latest optimized intermediate open-pit mine feature vector corresponding to the pending member is multiplied by the intermediate open-pit mine feature vector of the above-mentioned related next pending member in the previous optimization process, to obtain a corresponding similarity matrix, then the intermediate open-pit mine feature vector corresponding to the pending member is weighted based on the above-mentioned similarity matrix, to obtain a similarity vector, the similarity vector is added to the intermediate open-pit mine feature vector of the above-mentioned related next pending member in the previous optimization process, to obtain a corresponding associated optimization vector; after linear mapping of the associated optimization vector corresponding to each pending member related to the above-mentioned related next pending member, such as multiplication by a weight matrix, the weight matrix can be used as the network parameters of the corresponding neural network, and is constantly updated in the process of network training, based on the member association parameters corresponding to each pending member related to the above-mentioned related next pending member, the corresponding associated optimization vector is weighted and summed, to obtain the intermediate open-pit mine feature vector of the above-mentioned related next pending member after the current optimization, to optimize the intermediate open-pit mine feature vector in the previous optimization process.

[0051] In some embodiments, the step S2021 can be implemented by the following steps: a step S20211, obtaining distribution ordering information between the pending members, wherein the distribution ordering information is used to represent the distribution ordering between the pending members, the pending members corresponding to the same description data have the same distribution ordering, and the pending members corresponding to different description data have different member distribution orderings; a step S20212, determining a local area position corresponding to the pending members according to feature information corresponding to the pending members, wherein the area granularity corresponding to the feature information is greater than or equal to the area granularity of the local area position; a step S20213, performing a deduplication processing on the local area position to obtain at least one target local position, and marking the at least one target local position to obtain a link relationship member; a step S20214, determining member link information between the pending members according to the link relationship member, wherein the member link information is used to represent the link relationship between the pending members, the pending members having different distribution orderings are linked through at least one link relationship member, for any one of the pending members linked by any one of the link relationship members, the local area position corresponding to the feature information corresponding to the pending member includes the target local position corresponding to the link relationship member, and in the distribution ordering information, the link relationship members have the same distribution ordering; and a step S20215, forming a corresponding member relationship network according to each of the pending members, the distribution ordering information, and the member link information. The method can further increase the content of the member relationship network.

[0052] Exemplarily, the to-be-determined members corresponding to the description data of the image dimension have the same member distribution order, the to-be-determined members corresponding to the description data of the text dimension have the same member distribution order, and the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine and the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine have different member distribution orders. In addition, for the to-be-determined members corresponding to the description data of the open-pit mine of different dimensions, the specific member distribution order is not limited, and can be that the member distribution order of the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine is in the front, the member distribution order of the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine is in the back, or the member distribution order of the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine is in the back, and the member distribution order of the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine is in the front. In the above member relationship network, the member distribution order can have three kinds, such as a first kind of member distribution order, corresponding to the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine, a second kind of member distribution order, corresponding to each link relationship member, and a third kind of member distribution order, corresponding to the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine. The link relationship member is between the to-be-determined members, and can be used to reflect whether the to-be-determined members have relevance. That is, if there is a link relationship member between two to-be-determined members, it indicates that the two to-be-determined members have relevance, and if there is no link relationship member between two to-be-determined members, it indicates that the two to-be-determined members do not have relevance.

[0053] For example, the pending members for the image-level open-pit mine description data include pending member 1, pending member 2, and pending member 3. The feature data corresponding to pending member 1 is an image of the mining pit, the feature data corresponding to pending member 2 is an image of the mining pit, and the feature data corresponding to pending member 3 is an image of the mining pit. The pending members for the text-level open-pit mine description data include pending member 4, pending member 5, and pending member 6. The feature data corresponding to pending member 4 is the recorded text of the mining pit, the feature data corresponding to pending member 5 is the recorded text of the mining pit, and the feature data corresponding to pending member 6 is the recorded text of the mining pit. Based on the image corresponding to pending member 1, the determined local location includes the excavation area (including the specific area where mining activities are taking place) and the slope area (referring to the edge of the mining pit, which may contain high and low slopes). Based on the image corresponding to pending member 2, the determined local location includes the excavation area. Based on the image corresponding to pending member 3, the determined local location includes the excavation area and the bottom area; based on the recorded text corresponding to pending member 4, the determined local location includes the excavation area and the slope area. Based on the record text corresponding to undetermined member 5, the determined local locations include the slope area and the bottom area. Based on the record text corresponding to undetermined member 6, the determined local locations include the excavation area and the bottom area. Therefore, the determined important local locations can be the excavation area, the slope area, and the bottom area. Thus, three corresponding link network members can be formed: link member 1 (corresponding to the excavation area), link member 2 (corresponding to the slope area), and link member 3 (corresponding to the bottom area). Based on this, the following is obtained: Figure 3 The member relationship network shown has the following relationships: Members 1, 2, and 3 of the pending relationship network have the same member distribution order; members 4, 5, and 6 of the pending relationship network have the same member distribution order; and members 1, 2, and 3 of the linked relationship network have the same member distribution order. Member 1 of the linked relationship network can be linked to members 1, 2, 3, 4, and 6 of the pending relationship network; member 2 of the linked relationship network can be linked to members 1, 4, and 5 of the pending relationship network; and member 3 of the linked relationship network can be linked to members 3, 5, and 6 of the pending relationship network.

[0054] Following step S20215 and preceding step S2022, the method further includes step S20216, updating the distribution sorting of the undetermined members based on the aforementioned distribution sorting information. This method can further improve the reliability of data collection.

[0055] Specifically, the optimization order between the undetermined members corresponding to the two data dimensions can be determined arbitrarily. For example, in the first optimization process, the undetermined members corresponding to the image dimension can be optimized first, and the undetermined members corresponding to the text dimension can be optimized later. In the second optimization process, the undetermined members corresponding to the text dimension can be optimized first, and the undetermined members corresponding to the image dimension can be optimized later. In this way, the optimization process can be performed alternately. It should be noted that, in the optimization process, each related undetermined member of a certain undetermined member does not have the same order relationship with the certain undetermined member. The related undetermined member is the previous related undetermined member, and does not include the next related undetermined member when there are three or more levels. That is, in one optimization process, the information transmission of the undetermined members is unidirectional.

[0056] In another embodiment, the step S20222 further comprises: a step S202221 of obtaining a statistical distribution parameter of the target local position of the link relationship member; and a step S202222 of calculating a link edge association parameter according to the statistical distribution parameter to obtain the first association parameter, wherein the link edge association parameter is used to represent the link relationship between the link relationship member and the undetermined member. This method can further improve the accuracy of calculating the first association parameter.

[0057] Specifically, in the weighting process based on the member association parameter, the member association parameter can be updated based on the link edge association parameter, for example, by multiplication, and then weighting based on the updated member association parameter. For example, when updating the intermediate feature vector of the undetermined member 5 in the previous optimization process based on the undetermined member 1, the member association parameter between the undetermined member 1 and the undetermined member 5 needs to be updated based on the link edge association parameter between the undetermined member 1 and the link relationship member 2 and the link edge association parameter between the undetermined member 5 and the link relationship member 2, for example, by multiplication. For any link relationship network member and any undetermined member, in the regional local positions of the target open-pit mine area reflected by the any undetermined member, the statistical distribution parameter of the important regional local position corresponding to the any link relationship network member is determined, the regional local positions of the target open-pit mine area reflected by the any undetermined member are at least one, and the statistical distribution parameter can be the ratio between the number of important regional local positions and the number of each regional local position of the target open-pit mine area reflected by the any undetermined member. The link edge association parameter between the any link relationship network member and the any undetermined member is determined based on the statistical distribution parameter, and the link edge association parameter can be positively correlated with the statistical distribution parameter, and the statistical distribution parameter can be directly used as the link edge association parameter.

[0058] Step S204 can be implemented by the following steps, for example: step S2041, obtaining a plurality of above-mentioned target member information corresponding to the above-mentioned description data, obtaining a plurality of preliminary data; step S2042, combining a plurality of above-mentioned preliminary data to obtain the above-mentioned target data. This method can further improve the reliability of the above-mentioned target data by combining a plurality of preliminary data.

[0059] Specifically, the feature information corresponding to at least one target relationship network member corresponding to one kind of description data can be combined together to form an open pit representative data corresponding to one kind of open pit description data; the open pit representative data corresponding to each kind of open pit description data can be combined together to form the above-mentioned target data.

[0060] The embodiment of the present application also provides a data acquisition device based on a 3D map. It should be noted that the data acquisition device based on a 3D map of the embodiment of the present application can be used to execute the data acquisition method based on a 3D map provided by the embodiment of the present application. The device is used to realize the above-mentioned embodiment and preferred embodiment, and will not be described here. As used below, the term "module" can be a combination of software and / or hardware that realizes a predetermined function. Although the device described in the following embodiment is preferably realized in software, hardware, or a combination of software and hardware is also possible and is contemplated.

[0061] The following describes the data acquisition device based on a 3D map provided by the embodiment of the present application.

[0062] Figure 4 is a schematic diagram of the data acquisition device based on a 3D map according to the embodiment of the present application. As shown in Figure 4 , the device includes:

[0063] The first acquisition unit 10 is configured to acquire description data of a target region of an open pit and determine a plurality of undetermined member information corresponding to the description data according to the description data, wherein the description data is data describing information of the target region from one dimension, and one undetermined member information is used to represent one feature information of the description data.

[0064] Specifically, at least one pending member corresponding to each open-pit mine description data can be determined according to the determined plurality of open-pit mine description data. Each pending member is used to reflect a feature information included in the corresponding open-pit mine description data, that is, a key data in the corresponding open-pit mine description data. One open-pit mine description data can correspond to one or more key data, that is, one or more pending members. The plurality of open-pit mine description data is used to respectively characterize the regional information of the target open-pit mine area from multiple dimensions. That is, one open-pit mine description data is used to respectively characterize the regional information of the target open-pit mine area from one dimension. The dimension can refer to the way of forming the open-pit mine description data, or the dimension can also refer to the data form of the open-pit mine description data, such as image data, text data, etc. For example, one frame of image is obtained by image acquisition of one position in the target open-pit mine area. For another example, the key data of one open-pit mine description data is formed by recording the mining process of a sub-region of the target open-pit mine area.

[0065] The first determining unit 20 is configured to determine a target feature vector corresponding to each pending member according to a correlation between any two of the plurality of pending member information, wherein the target feature vector is used to represent a vectorization result of the feature information.

[0066] Specifically, the feature information reflected by each pending member can be vectorized, and then the vectorization result can be associated and mined based on the correlation between the pending members to optimize the initial vector and obtain the target feature vector corresponding to each pending member, so that the representation ability of the target feature vector is better. For example, the pending member in the image dimension learns the feature information of the pending member in the text dimension.

[0067] The second determining unit 30 is configured to perform clustering processing on the target feature vector to obtain at least one cluster center vector, and determine the pending member information corresponding to the target feature vector as target member information.

[0068] Specifically, whether the corresponding target feature vector is representative or whether it meets other conditions can be used to determine whether the pending relationship network member belongs to the target member. Through the above-mentioned manner, at least one target member, such as image A or text segment 1, can be determined.

[0069] The second obtaining unit 40 is configured to obtain the description data corresponding to the target member information to obtain target data.

[0070] Specifically, the target data is used to update the 3D map of the open-pit mine constructed for the target open-pit mine area. The specific map updating method can adopt the prior art. The application focuses on how to collect and screen data, and therefore is not specifically limited in the application. In some embodiments, the target data can also be used as a basis for map updating. For example, similarity calculation can be performed based on the current target open-pit mine area data and the historical target open-pit mine area data. The similarity calculation can be any data similarity calculation, such as text similarity and image similarity. The similarities of different dimensions of data are fused, such as mean value. When the obtained data similarity is less than a preset similarity, it can be determined that the mining area has changed greatly and map updating is needed. When the obtained data similarity is greater than or equal to the preset similarity, it can be determined that the mining area has changed slightly and map updating is not needed.

[0071] Suppose there are three types of description data: image sequence A, image sequence B and text data 1. The determined pending members are: the pending member corresponding to image sequence A, pending member 1 and pending member 2; the pending member corresponding to image sequence B: pending member 3 and pending member 4; the pending member corresponding to text data 1: pending member 5 and pending member 6. The determined corresponding member correlation of each pending member can be: there is a correlation between pending member 1 and pending member 3; there is a correlation between pending member 2 and pending member 4; there is a correlation between pending member 3 and pending member 6; there is a correlation between pending member 4 and pending member 5. The target feature vector corresponding to pending member 1 is feature vector 1, the target feature vector corresponding to pending member 2 is feature vector 2, the target feature vector corresponding to pending member 3 is feature vector 3, the target feature vector corresponding to pending member 4 is feature vector 4, and the target feature vector corresponding to pending member 5 is feature vector 5. The target feature vector corresponding to pending member 6 is feature vector 6. Suppose the pending members corresponding to feature vector 1 and feature vector 2 are determined as target members, and the target data can be determined based on the pending members corresponding to the two feature vectors.

[0072] In the embodiment, the first acquisition unit acquires the description data of the target region of the open-pit mine, and determines a plurality of pending member information corresponding to the description data according to the description data. The description data is data for describing information of the target region from one dimension, and one pending member information is used to represent one characteristic information of the description data. The first determination unit determines a target feature vector corresponding to the pending member according to a correlation between any two pending members in the plurality of pending member information. The target feature vector is used to represent a vectorization result of the characteristic information. The second determination unit performs clustering processing on the target feature vector to obtain at least one cluster center vector, and determines that the pending member information corresponding to the target feature vector is target member information. The second acquisition unit acquires description data corresponding to the target member information to obtain target data. In the above scheme, in the process of screening the target open-pit mine region data from a plurality of open-pit mine description data, one open-pit mine description data is refined into at least one corresponding pending member, that is, one characteristic information, so that the granularity of the analysis object is smaller, and the data acquisition accuracy is further improved. Therefore, the data acquisition accuracy for constructing the 3D map of the open-pit mine can be effectively improved. The problem of low accuracy of the data acquisition method based on the 3D map in the prior art is solved.

[0073] In the specific implementation process, the description data is image data or text data. When the description data is the image data, the first acquisition unit includes a first acquisition module, a first determination module, a second acquisition module, and a second determination module. The first acquisition module is configured to acquire at least one description image in the description data. The description image is obtained by image information acquisition on the target region. The first determination module is configured to acquire first characteristic information corresponding to the description image, and determine the pending member information corresponding to the description data according to the first characteristic information. When the description data is the text data, the second acquisition module is configured to acquire at least one text segment in the description data. The text segment is obtained by text recording on the mining process in the target region. The second determination module is configured to mark the text segment as second characteristic information, and determine the pending member information corresponding to the description data according to the second characteristic information. The device can further quickly determine the plurality of pending member information corresponding to the description data.

[0074] Specifically, in the case that the description data is the image data, in the one image sequence, each open-pit mine description image, such as an image of a mining pit or an image of one angle in a smaller mining pit, can be extracted. Each open-pit mine description image is taken as one kind of feature information, and 3D image acquisition (i.e., a depth image) can be performed by a flying device, for example. In the case that the description data is the text data, in the one description data, at least one data segment, such as data of one region, is determined, and the data segment is formed by recording the mining process in the target region to obtain recorded text of each region.

[0075] To further improve the reliability of data acquisition, the first determining unit includes a generating module and a third determining module. The generating module is configured to generate a member relationship network according to the correlation between any two of the plurality of pending members. The third determining module is configured to determine the target feature vector corresponding to the pending member according to the member relationship network and a cyclic optimization rule. The device can cyclically update the correlation between the pending members by generating the member relationship network and combining the cyclic optimization rule, thereby learning the information of the pending member in another dimension, and the representation ability of the target feature vector can be enhanced. Therefore, the reliability of data acquisition for open-pit mine 3D map construction can be effectively improved.

[0076] Specifically, the member relationship network can be used to reflect the correlation between the determined pending members.

[0077] The third determining module includes a third obtaining module, a calculating module, a first processing module, and a repeating module. The third obtaining module is configured to obtain a first initial feature vector of a neighboring member of the pending member in a step, wherein the neighboring member is linked to the pending member, and the distribution order of the neighboring member is after the distribution order of the pending member. The calculating module is configured to calculate an association parameter between the pending member and the neighboring member according to the first initial feature vector in a calculating step, to obtain a first association parameter. The first processing module is configured to optimize the first initial feature vector according to the first association parameter in a processing step, to obtain an intermediate feature vector. The repeating module is configured to repeat the obtaining step, the calculating step, and the processing step at least once in a repeating step, and update the first initial feature vector in the obtaining step to the intermediate feature vector in the processing step obtained in the last repeating process in the repeating process. The intermediate feature vector in the case that a preset condition is reached is determined as the target feature vector, wherein the preset condition is that a preset number of times is reached. The device can further compensate for the limitation of information in different dimensions in representing information.

[0078] Specifically, the initial feature vector of the undetermined network member can be obtained by performing convolution operation on the feature information corresponding to the undetermined member, wherein the feature information of the image dimension and the feature information of the text dimension can be realized by different neural networks, such as different convolution neural networks, and the specific implementation is not limited, and the existing processing process for feature extraction or feature mining of images and texts can be adopted.

[0079] For example, for the first optimization: the initial open pit feature vector of the undetermined member 1 is taken as the intermediate open pit feature vector after the first optimization; the initial open pit feature vector of the undetermined member 2 is taken as the intermediate open pit feature vector after the first optimization; the initial open pit feature vector of the undetermined member 3 is taken as the intermediate open pit feature vector after the first optimization; for the undetermined member 4, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1, the undetermined member 2 and the undetermined member 3), the initial open pit feature vector of the undetermined member 4 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 4; for the undetermined member 5, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1 and the undetermined member 3), the initial open pit feature vector of the undetermined member 5 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 5; for the undetermined member 6, based on the intermediate open pit feature vectors of the related undetermined members (i.e. the intermediate open pit feature vectors after the first optimization corresponding to the undetermined member 1, the undetermined member 2 and the undetermined member 3), the initial open pit feature vector of the undetermined member 6 is optimized to form the intermediate open pit feature vector after the first optimization of the undetermined member 6.

[0080] For the second optimization: the intermediate open-pit mine feature vector of the first optimization of the pending member 4 is taken as the intermediate open-pit mine feature vector after the second optimization; the intermediate open-pit mine feature vector of the first optimization of the pending member 5 is taken as the intermediate open-pit mine feature vector after the second optimization; the intermediate open-pit mine feature vector of the first optimization of the pending member 6 is taken as the intermediate open-pit mine feature vector after the second optimization; for the pending member 1, the intermediate open-pit mine feature vector of the first optimization of the pending member 1 is optimized based on the intermediate open-pit mine feature vectors of the related pending members (i.e., the intermediate open-pit mine feature vectors of the first optimization corresponding to the pending member 4, the pending member 5 and the pending member 6) to form the intermediate open-pit mine feature vector of the second optimization of the pending member 1; for the pending member 2, the intermediate open-pit mine feature vector of the first optimization of the pending member 2 is optimized based on the intermediate open-pit mine feature vectors of the related pending members (i.e., the intermediate open-pit mine feature vectors of the first optimization corresponding to the pending member 4 and the pending member 6) to form the intermediate open-pit mine feature vector of the second optimization of the pending member 2; for the pending member 3, the intermediate open-pit mine feature vector of the first optimization of the pending member 3 is optimized based on the intermediate open-pit mine feature vectors of the related pending members (i.e., the intermediate open-pit mine feature vectors of the first optimization corresponding to the pending member 4 and the pending member 6) to form the intermediate open-pit mine feature vector of the second optimization of the pending member 3. Through the continuous circulation of the above optimization steps, the optimization of the two mutually inverse sequential relationships from top to bottom and from bottom to top can be realized, and through the successive circulation, the feature information between the pending members in the above member relationship network can be fully interacted, and the number of specific circulation is not limited.

[0081] The first processing module includes a first acquisition submodule, a first calculation submodule and a first processing submodule. The first acquisition submodule is configured to acquire a second initial feature vector of another pending member of the pending member. The another pending member is in the same member relationship network as the pending member and is different from the pending member. The first calculation submodule is configured to calculate a second correlation parameter of the pending member and the adjacent member according to the initial feature vector, to obtain the second correlation parameter. The first processing submodule is configured to perform weighted processing on the second initial feature vector according to the second correlation parameter, to obtain the intermediate feature vector. The device can further quickly calculate the intermediate feature vector.

[0082] Specifically, for each pending member related to the above-mentioned related next pending member, including the above-mentioned one pending member, at least one other member related to the above-mentioned related next pending member, the latest optimized intermediate open-pit mine feature vector corresponding to the pending member is multiplied by the intermediate open-pit mine feature vector of the above-mentioned related next pending member in the previous optimization process, to obtain a corresponding similarity matrix, then the intermediate open-pit mine feature vector corresponding to the pending member is weighted based on the above-mentioned similarity matrix, to obtain a similarity vector, the similarity vector is added to the intermediate open-pit mine feature vector of the above-mentioned related next pending member in the previous optimization process, to obtain a corresponding associated optimization vector; after linear mapping of the associated optimization vector corresponding to each pending member related to the above-mentioned related next pending member, such as multiplication by a weight matrix, the weight matrix can be used as the network parameters of the corresponding neural network, and is constantly updated in the process of network training, based on the member association parameters corresponding to each pending member related to the above-mentioned related next pending member, the corresponding associated optimization vector is weighted and summed, to obtain the intermediate open-pit mine feature vector of the above-mentioned related next pending member after the current optimization, to optimize the intermediate open-pit mine feature vector in the previous optimization process.

[0083] In some embodiments, the generating module comprises a second obtaining sub-module, a first determining sub-module, a second processing sub-module, a second determining sub-module, and a forming sub-module. The second obtaining sub-module is configured to obtain distribution ordering information between the to-be-determined members, wherein the distribution ordering information is used to represent the distribution ordering between the to-be-determined members, the to-be-determined members corresponding to the same description data have the same distribution ordering, and the to-be-determined members corresponding to different description data have different member distribution orderings. The first determining sub-module is configured to determine a local area position corresponding to each to-be-determined member according to feature information corresponding to the to-be-determined member, wherein the area granularity corresponding to the feature information is greater than or equal to the area granularity of the local area position. The second processing sub-module is configured to perform deduplication processing on the local area position to obtain at least one target local position, and mark the at least one target local position to obtain a link relationship member. The second determining sub-module is configured to determine member link information between the to-be-determined members according to the link relationship member, wherein the member link information is used to represent the link relationship between the to-be-determined members, the to-be-determined members having different distribution orderings are linked through at least one link relationship member, for any to-be-determined member linked by any link relationship member, the local area position corresponding to the feature information corresponding to the to-be-determined member includes a target local position corresponding to the link relationship member, and in the distribution ordering information, the link relationship members have the same distribution ordering. The forming sub-module is configured to form a corresponding member relationship network according to each to-be-determined member, the distribution ordering information, and the member link information. The device can further increase the content of the member relationship network.

[0084] Exemplarily, the to-be-determined members corresponding to the description data of the image dimension have the same member distribution order, the to-be-determined members corresponding to the description data of the text dimension have the same member distribution order, and the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine and the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine have different member distribution orders. In addition, for the to-be-determined members corresponding to the description data of the open-pit mine of different dimensions, the specific member distribution order is not limited, and can be that the member distribution order of the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine is in the front, the member distribution order of the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine is in the back, or the member distribution order of the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine is in the back, and the member distribution order of the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine is in the front. In the above member relationship network, the member distribution order can have three kinds, such as a first kind of member distribution order, corresponding to the to-be-determined members corresponding to the description data of the image dimension of the open-pit mine, a second kind of member distribution order, corresponding to each link relationship member, and a third kind of member distribution order, corresponding to the to-be-determined members corresponding to the description data of the text dimension of the open-pit mine. The link relationship member is between the to-be-determined members, and can be used to reflect whether the to-be-determined members have relevance. That is, if there is a link relationship member between two to-be-determined members, it indicates that the two to-be-determined members have relevance, and if there is no link relationship member between two to-be-determined members, it indicates that the two to-be-determined members do not have relevance.

[0085] For example, the open-pit mine description data corresponding to the image dimension pending member includes pending member 1, pending member 2 and pending member 3, the feature data corresponding to the pending member 1 is the image of the mining pit, the feature data corresponding to the pending member 2 is the image of the mining pit, and the feature data corresponding to the pending member 3 is the image of the mining pit; the open-pit mine description data corresponding to the text dimension pending member includes pending member 4, pending member 5 and pending member 6, the feature data corresponding to the pending member 4 is the recorded text of the mining pit, the feature data corresponding to the pending member 5 is the recorded text of the mining pit, and the feature data corresponding to the pending member 6 is the recorded text of the mining pit; based on the image corresponding to the pending member 1, the determined regional local position includes the excavation area (including the specific area where the mining activity is being carried out), the slope area (referring to the edge part of the mining pit, which may have high and low slopes). Based on the image corresponding to the pending member 2, the determined regional local position includes the excavation area. Based on the image corresponding to the pending member 3, the determined regional local position includes the excavation area and the bottom area; based on the recorded text corresponding to the pending member 4, the determined regional local position includes the excavation area and the slope area. Based on the recorded text corresponding to the pending member 5, the determined regional local position includes the slope area and the bottom area. Based on the recorded text corresponding to the pending member 6, the determined regional local position includes the excavation area and the bottom area. Based on this, the important regional local position determined can be the excavation area, the slope area and the bottom area, so that the corresponding three link relationship network members can be formed, which are link relationship member 1 (corresponding to the excavation area), link relationship member 2 (corresponding to the slope area) and link relationship member 3 (corresponding to the bottom area); based on this, the member relationship network as shown in FIG. 8 is obtained, the pending relationship network member 1, the pending relationship network member 2 and the pending relationship network member 3 have the same member distribution order, the pending relationship network member 4, the pending relationship network member 5 and the pending relationship network member 6 have the same member distribution order, and the link relationship network member 1, the link relationship network member 2 and the link relationship network member 3 have the same member distribution order. The link relationship network member 1 can have a link relationship with the pending relationship network member 1, the pending relationship network member 2, the pending relationship network member 3, the pending relationship network member 4 and the pending relationship network member 6 respectively; the link relationship network member 2 can have a link relationship with the pending relationship network member 1, the pending relationship network member 4 and the pending relationship network member 5 respectively; and the link relationship network member 3 can have a link relationship with the pending relationship network member 3, the pending relationship network member 5 and the pending relationship network member 6 respectively. Figure 3

[0086] The device further includes an updating submodule for updating the distribution order of the pending member according to the distribution order information. The device can further improve the reliability of data acquisition.

[0087] ​Specifically, the optimization order between the undetermined members corresponding to the two data dimensions can be determined arbitrarily. For example, in the first optimization process, the undetermined members corresponding to the image dimension can be optimized first, and the undetermined members corresponding to the text dimension can be optimized later. In the second optimization process, the undetermined members corresponding to the text dimension can be optimized first, and the undetermined members corresponding to the image dimension can be optimized later. In this way, the optimization process can be performed alternately. It should be noted that, in the optimization process, each related undetermined member of a certain undetermined member does not have the same order with the certain undetermined member. The related undetermined member is the previous level, and does not include the next level when there are three or more levels. That is, in one optimization process, the information transmission of the undetermined members is unidirectional.

[0088] In another embodiment, the computing module includes a third acquisition submodule and a second calculation submodule. The third acquisition submodule is configured to acquire a statistical distribution parameter of the target local position of the link relationship member. The second calculation submodule is configured to calculate a link edge association parameter based on the statistical distribution parameter to obtain the first association parameter, wherein the link edge association parameter is used to represent the link relationship between the link relationship member and the undetermined member. The device can further improve the accuracy of calculating the first association parameter.

[0089] Specifically, in the weighting process based on the member association parameter, the member association parameter can be updated based on the link edge association parameter, for example, by multiplying, and then weighting based on the updated member association parameter. For example, when updating the intermediate feature vector of the undetermined member 5 in the previous optimization process based on the undetermined member 1, the member association parameter between the undetermined member 1 and the undetermined member 5 needs to be updated based on the link edge association parameter between the undetermined member 1 and the link relationship member 2 and the link edge association parameter between the undetermined member 5 and the link relationship member 2, such as multiplication. For any one link relationship network member and any one undetermined member, in the regional local positions of the target open-pit mine area reflected by the any one undetermined member, the statistical distribution parameter of the important regional local position corresponding to the any one link relationship network member is determined. The regional local position of the target open-pit mine area reflected by the any one undetermined member is at least one, and the statistical distribution parameter can be the ratio between the number of important regional local positions and the number of regional local positions of the target open-pit mine area reflected by the any one undetermined member. The link edge association parameter between the any one link relationship network member and the any one undetermined member is determined based on the statistical distribution parameter. The link edge association parameter can be positively correlated with the statistical distribution parameter, and the statistical distribution parameter can be directly used as the link edge association parameter.

[0090] The second obtaining unit includes a fourth obtaining module and a second processing module, the fourth obtaining module is configured to obtain the description data corresponding to the plurality of target member information, and obtain a plurality of preliminary data; and the second processing module is configured to combine the plurality of preliminary data to obtain the target data.

[0091] Specifically, the feature information corresponding to at least one target relationship network member corresponding to one kind of description data can be combined together to form open pit mine representative data corresponding to one kind of open pit mine description data; and the open pit mine representative data corresponding to each kind of open pit mine description data can be combined together to form the target data.

[0092] The data acquisition device based on the 3D map includes a processor and a memory, the first obtaining unit, the first determining unit, the second determining unit and the second obtaining unit are stored in the memory as program units, and the corresponding functions are realized by the processor executing the program units stored in the memory. The modules are located in the same processor; or the modules are located in different processors in any combination.

[0093] The processor includes a core, and the core retrieves the corresponding program unit from the memory. The core can be one or more, and the data is collected by adjusting the core parameters.

[0094] The memory can include a non-permanent memory in a computer readable medium, a random access memory (RAM) and / or a non-volatile memory such as a read-only memory (ROM) or a flash memory (flash RAM), and the memory includes at least one memory chip.

[0095] The embodiment of the present application provides a computer readable storage medium, the computer readable storage medium includes a stored program, wherein the computer readable storage medium controls the device to execute the data acquisition method based on the 3D map when the program runs.

[0096] The embodiment of the present application provides a processor, the processor is used for running a program, wherein the processor executes the data acquisition method based on the 3D map when the program runs.

[0097] The embodiment of the present application provides a device, the device includes a processor, a memory and a program stored in the memory and executable on the processor, and the processor executes the data acquisition method based on the 3D map when the program runs. The device in the present application can be a server, a PC, a PAD, a mobile phone and the like.

[0098] The application also provides a computer program product adapted to execute, when executed on a data processing device, a program initializing the steps of the above-mentioned data collection method based on a 3D map.

[0099] It is apparent that those skilled in the art should understand that the above-mentioned modules or steps of the application can be realized by general computing devices, which can be centralized on a single computing device or distributed on a network composed of multiple computing devices, and can be realized by program codes executable by the computing devices, so that they can be stored in storage devices and executed by the computing devices, and in some cases, the steps shown or described can be executed in different orders, or they can be respectively manufactured into individual integrated circuit modules, or multiple modules or steps among them can be manufactured into a single integrated circuit module. Thus, the application is not limited to any specific combination of hardware and software.

[0100] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer-usable program code.

[0101] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks

[0102] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction devices that implement the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The functions specified in a flow or multiple flows and / or blocks

[0103] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 Figure 1

[0104] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0105] The memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) about which the processor can execute instructions. The memory can also include non-volatile memory, such as read only memory (ROM), electrically programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), erasable programmable read only memory (EPROM), flash memory, or a combination of non-volatile memories in different types. The memory can also include a compact disk read only memory (CD-ROM), digital versatile disk (DVD), Blu-ray, or another non-transitory computer readable medium, which is non-volatile and non-transitory in nature, but volatile in that it can lose its content if the power to the computer is turned off or if the computer crashes. The memory is an example of a computer readable medium.

[0106] Computer readable media includes permanent and non-permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory (ROM), electrically programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD), or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer readable media does not include transitory media such as modulated data signals and carriers.

[0107] It should also be noted that the terms "comprising," "including," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements recited, but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0108] ​​The above descriptions are only the preferred embodiments of the present application, and are not intended to limit the present application. The present application can have various modifications and changes for those skilled in the art. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A data acquisition method based on 3D maps, characterized in that, include: Obtain descriptive data of the target area of ​​the open-pit mine, and determine multiple pending member information corresponding to the descriptive data based on the descriptive data. The descriptive data is data that describes the information of the target area from one dimension, and each pending member information is used to characterize a feature information of the descriptive data. Based on the correlation between any two of the undetermined members among the multiple undetermined member information, a target feature vector corresponding to the undetermined member is determined, wherein the target feature vector is used to characterize the vectorization result of the feature information; Clustering is performed on the target feature vector to obtain at least one cluster center vector, and the undetermined member information corresponding to the target feature vector is determined as the target member information; Obtain the description data corresponding to the target member information to obtain the target data. The description data is image data or text data. When the description data is image data, multiple undetermined member information corresponding to the description data is determined based on the description data, including: At least one descriptive image is obtained from the descriptive data, wherein the descriptive image is obtained by acquiring image information of the target region; Obtain the first feature information corresponding to the description image, and determine the undetermined member information corresponding to the description data based on the first feature information; When the description data is the text data, based on the description data, determine multiple undetermined member information corresponding to the description data, including: At least one text fragment is obtained from the description data, wherein the text fragment is obtained by textually recording the mining process in the target area; The text fragment is marked as second feature information, and the undetermined member information corresponding to the description data is determined based on the second feature information.

2. The method according to claim 1, characterized in that, Based on the correlation between any two of the undetermined members from a plurality of undetermined member information, the target feature vector corresponding to the undetermined member is determined, including: Generate a member relationship network based on the correlation between any two of the multiple undetermined members; Based on the member relationship network and the cyclic optimization rule, the target feature vector corresponding to the undetermined member is determined.

3. The method according to claim 2, characterized in that, Based on the member relationship network and the iterative optimization rule, the target feature vector corresponding to the undetermined member is determined, including: The acquisition step involves acquiring the first initial feature vector of the neighboring members of the member to be determined, wherein the neighboring members are linked to the member to be determined, and the distribution order of the neighboring members follows the distribution order of the member to be determined. The calculation steps are as follows: based on the first initial feature vector, calculate the association parameters between the member to be determined and the adjacent members to obtain the first association parameter; The processing step involves optimizing the first initial feature vector based on the first association parameter to obtain an intermediate feature vector. The steps are repeated, including the acquisition step, the calculation step, and the processing step, at least once. During the repetition, the first initial feature vector in the acquisition step is updated to the intermediate feature vector obtained in the processing step in the previous repetition. The intermediate feature vector that meets a preset condition is determined as the target feature vector, wherein the preset condition is meeting a preset number of conditions.

4. The method according to claim 3, characterized in that, The processing steps include: Obtain the second initial feature vectors of other members to be determined, wherein the other members to be determined are located in the same member relationship network as the member to be determined, and the other members to be determined are different from the member to be determined; Based on the initial feature vector, the association parameters between the undetermined member and the adjacent members are calculated to obtain the second association parameter; The second initial feature vector is weighted according to the second association parameter to obtain the intermediate feature vector.

5. The method according to claim 3, characterized in that, Based on the correlation between any two of the multiple undetermined members, a member relationship network is generated, including: Obtain the distribution and sorting information among the members to be determined, wherein the distribution and sorting information is used to characterize the distribution and sorting among the members to be determined, the members to be determined corresponding to the same description data have the same distribution and sorting, and the members to be determined corresponding to different description data have different member distribution and sorting. Based on the feature information corresponding to the undetermined member, the local region position corresponding to the undetermined member is determined, wherein the region granularity corresponding to the feature information is greater than or equal to the region granularity of the local region position. The local region locations are deduplicated to obtain at least one target local location, and the at least one target local location is marked to obtain the link relationship members; Based on the link relationship members, member link information between the undetermined members is determined, wherein the member link information is used to characterize the link relationship between the undetermined members, the undetermined members with different distribution orders are linked through at least one of the link relationship members, for any undetermined member linked by any link relationship member, the local location of the region corresponding to the feature information of the undetermined member includes the target local location corresponding to the link relationship member, and in the distribution order information, the link relationship members have the same distribution order; Based on each of the undetermined members, the distribution and sorting information, and the member link information, a corresponding member relationship network is formed.

6. The method according to claim 5, characterized in that, After forming a corresponding member relationship network based on each of the undetermined members, the distribution sorting information, and the member link information, and before determining the target feature vector corresponding to each undetermined member based on the member relationship network and the cyclic optimization rule, the method further includes: The distribution sorting of the undetermined members is updated based on the distribution sorting information.

7. The method according to claim 5, characterized in that, The calculation steps also include: Obtain the statistical distribution parameters of the target local locations of the linked members; Based on the statistical distribution parameters, the link edge association parameters are calculated to obtain the first association parameter, wherein the link edge association parameters are used to characterize the link relationship between the link relationship member and the undetermined member.

8. The method according to claim 1, characterized in that, Obtaining the description data corresponding to the target member information to obtain target data includes: Obtain the description data corresponding to multiple target member information to obtain multiple preliminary data; The target data is obtained by combining and processing multiple sets of the prepared data.

9. A data acquisition device based on a 3D map, characterized in that, include: The first acquisition unit is used to acquire descriptive data of the target area of ​​the open-pit mine, and determine multiple pending member information corresponding to the descriptive data based on the descriptive data. The descriptive data is data that describes the information of the target area from one dimension, and one of the pending member information is used to characterize a feature information of the descriptive data. The first determining unit is configured to determine the target feature vector corresponding to the undetermined member based on the correlation between any two undetermined members among the plurality of undetermined member information, wherein the target feature vector is used to characterize the vectorization result of the feature information; The second determining unit is used to perform clustering processing on the target feature vector to obtain at least one cluster center vector, and to determine the undetermined member information corresponding to the target feature vector as target member information; The second acquisition unit is used to acquire the description data corresponding to the target member information to obtain the target data. The description data is image data or text data. When the description data is image data, the first acquisition unit includes a first acquisition module, a first determination module, a second acquisition module, and a second determination module. The first acquisition module acquires at least one description image from the description data, wherein the description image is obtained by collecting image information from the target area. The first determination module acquires first feature information corresponding to the description image and determines the pending member information corresponding to the description data based on the first feature information. When the description data is text data, the second acquisition module acquires at least one text fragment from the description data, wherein the text fragment is obtained by recording the mining process in the target area. The second determination module marks the text fragment as second feature information and determines the pending member information corresponding to the description data based on the second feature information.

Images