CN-122024110-A - Unmanned aerial vehicle aerial photographing topographic mapping processing method, system, equipment and medium

CN122024110ACN 122024110 ACN122024110 ACN 122024110ACN-122024110-A

Abstract

The application relates to the technical field of topographic mapping, in particular to a topographic mapping processing method, system, equipment and medium for unmanned aerial vehicle aerial photography, which comprises the steps of obtaining current observation data of a region to be mapped of a plurality of unmanned aerial vehicles; the method comprises the steps of generating individual cognitive entangled states of local topography based on current observation data, carrying out consensus evolution on each individual cognitive entangled state of the same local topography to generate a group cognitive field, wherein the group cognitive field is used for representing a geometric form distribution field of the local topography, each unmanned plane evaluates cognitive conflict level based on the current state of the group cognitive field to obtain an expected excitation value, and determining an observation target and an observation path based on the expected excitation value to carry out topography mapping according to the observation target and the observation path. The sampling efficiency can be improved, and meanwhile, the topographic mapping precision is improved.

Inventors

YANG KEMING
Li pengda
YANG NING
LI XIAOLONG
XU JIFENG
ZHANG JINAI
YIN LIZHI
LV FENGLIN
QI YANPENG
JIA XIAOMING
HUO ZHIJUN
BAI YUNHAO
WANG JIANGJIANG
ZHANG SHANGWEI
ZHOU LIANG

Assignees

天津中腾测绘科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260224

Claims (10)

1. The unmanned aerial vehicle aerial photographing topographic mapping processing method is characterized by comprising the following steps of: acquiring current observation data of a region to be mapped of a plurality of unmanned aerial vehicles; Generating an individual cognitive entangled state of the local terrain based on the current observation data, wherein the individual cognitive entangled state comprises geometric virtual forms of at least two alternative terrains and corresponding confidence degrees, and the individual cognitive entangled state is used for representing quantum mixed state representation of each unmanned aerial vehicle on the local terrain; performing consensus evolution on each individual cognitive entangled state aiming at the same local topography to generate a group cognitive field, wherein the group cognitive field is used for representing a geometric distribution field of the local topography; each unmanned aerial vehicle evaluates the cognitive conflict level based on the current state of the group cognitive field to obtain an expected excitation value; based on the expected excitation values, an observation target and an observation path are determined for topographic mapping from the observation target and the observation path.
2. The method of claim 1, wherein the generating individual cognitive entangled states for local terrain based on the current observation data comprises: performing coherent computation on the current observation data, and extracting topological features of the local topography; Inputting the topological feature into a preset geometric virtual form generating network, converting the topological feature into the geometric virtual form by a tensor network in the preset geometric virtual form generating network, and outputting the geometric virtual form and confidence, wherein the confidence is used for representing entanglement entropy of the geometric virtual form.
3. The method of claim 2, wherein prior to said inputting the topological feature into a preset geometric virtual form generation network, the method further comprises: Generating an countermeasure network by utilizing a preset condition based on the current observation data and the corresponding fact observation condition vector, and generating a counterfact observation condition vector, wherein the counterfact observation condition vector comprises at least one of disturbance factors, geometric shielding scenes and sensor characteristic observation conditions; Inputting the current observation data, the fact observation condition vector and the corresponding anti-fact observation condition into a preset geometric virtual form generation network to generate a group of fact assumption and a plurality of groups of anti-fact assumption; Determining a contrast loss function based on the fact assumption and the inverse fact assumption; And under the condition that the contrast loss function does not meet the training stop condition, updating network parameters through a preset gradient descent strategy until the preset geometric virtual form generating network meets the training condition.
4. The method of claim 1, wherein said co-discrimination evolving each of said individual cognitive entangled states to generate a group cognitive field comprises: dispersing the region to be painted into a plurality of grid cells; Constructing an atomic geometric proposition set for each grid cell, wherein the atomic geometric proposition set comprises a plurality of propositions, and the propositions are used for describing the geometric states of the grid cells; for each unmanned aerial vehicle, converting each individual cognitive entangled state into a judgment set; Integrating each judgment set to obtain a candidate collective judgment set; For each judgment set, calculating the similarity between the candidate collective judgment set and the judgment set to obtain the overall similarity of each candidate collective judgment set; Constraint optimization solution is carried out on the overall similarity by using a preset certificate planning solution algorithm, so that an optimal collective judgment combination is obtained; and generating a group cognitive field based on the collective judgment set.
5. The method of claim 1, wherein each unmanned aerial vehicle evaluates a cognitive conflict level based on a current state of the group cognitive field, resulting in an expected incentive value, comprising: modeling the current state of the group cognitive field as a Markov random field, wherein the Markov random field comprises a plurality of nodes and edges, the nodes are used for representing probability distribution of each current state, and the edges are used for representing the geometrical and semantic coupling relation of different regional terrains; calculating free energy under the approximation of the Markov random field; and determining a predicted excitation value corresponding to the free energy based on the corresponding relation between the preset free energy and the excitation value.
6. The method of claim 5, wherein the interaction potential function of the markov random field is learned online from historical mapping data via a graph neural network.
7. The method of claim 1, wherein the determining an observation target and an observation path based on the expected excitation value comprises: based on the current state of the group cognitive field, calculating to obtain a space-time four-dimensional excitation field covering the region to be painted; In the space-time four-dimensional excitation field, determining a space-time track of each unmanned aerial vehicle in a preset future time period as a decision variable, and carrying out optimization solution according to a preset constraint set and the decision variable by using a preset distributed constraint optimization algorithm to obtain the observation target and the observation path.
8. A terrain mapping processing system for unmanned aerial vehicle aerial photography, the system comprising: the acquisition module is used for acquiring current observation data of the region to be mapped of the unmanned aerial vehicle; the generation module is used for generating an individual cognition entanglement state of the local terrain based on the current observation data, wherein the individual cognition entanglement state comprises geometric virtual forms of at least two alternative terrains and corresponding confidence degrees, and the individual cognition entanglement state is used for representing two-word mixed state representation of each unmanned plane on the local terrain; The evolution module is used for carrying out consensus evolution on each individual cognition entanglement state aiming at the same local topography to generate a group cognition field, wherein the group cognition field is used for representing a geometric form distribution field of the local topography; The evaluation module is used for evaluating the cognitive conflict level by each unmanned aerial vehicle based on the current state of the group cognitive field to obtain an expected excitation value; And the determining module is used for determining an observation target and an observation path based on the expected excitation value so as to conduct topographic mapping according to the observation target and the observation path.
9. An electronic device comprising a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements the unmanned aerial vehicle topographic mapping method according to any one of claims 1-7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon computer program instructions, which when executed by a processor, implement a method of unmanned aerial vehicle aerial topographic mapping according to any of claims 1-7.

Description

Unmanned aerial vehicle aerial photographing topographic mapping processing method, system, equipment and medium Technical Field The application relates to the technical field of unmanned aerial vehicle mapping, in particular to a method, a system, equipment and a medium for processing unmanned aerial vehicle aerial photo topographic mapping. Background At present, along with the development of unmanned aerial vehicle technology, the intelligent information acquisition efficiency is improved in all fields, particularly in the field of topographic mapping, the aerial photogrammetry is rapidly developed, large-range topographic data is allowed to be acquired from the air, the mapping efficiency is greatly improved, after the unmanned aerial vehicle enters the 21 st century, the unmanned aerial vehicle becomes an important tool for topographic mapping, the unmanned aerial vehicle carries various sensors, the topographic data acquisition can be carried out in a smaller range with higher resolution, the development of the topographic mapping processing method is the result of multidisciplinary intersection, and the unmanned aerial vehicle relates to a plurality of fields such as mathematics, physics, engineering, computer science and the like. In the related art, a standard flow of topographic mapping by using an unmanned aerial vehicle is that a fixed flight route (such as a grid-shaped route) is designed in advance based on a task area boundary and a desired ground sampling distance, the unmanned aerial vehicle collects all images according to a plan, and finally, intensive calculation is carried out in a background workstation to generate a digital surface model or a three-dimensional real-scene model. In the related art, the unmanned aerial vehicle is used for carrying out topographic mapping, so that the problem that sampling is insufficient for a simple flat area and a complex topographic or texture missing area possibly exists in China, so that later modeling fails or the precision is uneven is solved, meanwhile, the data defect cannot be found and remedied in the acquisition stage, only a loophole can be found manually in the later processing, and the efficiency is low and the precision is low. Disclosure of Invention The embodiment of the application provides a method, a system, equipment and a medium for processing the topographic mapping of unmanned aerial vehicle aerial photography, which can improve the sampling efficiency and the topographic mapping precision. In one aspect, an embodiment of the present application provides a method for mapping and processing a topography of an unmanned aerial vehicle, where the method includes: acquiring current observation data of a region to be mapped of a plurality of unmanned aerial vehicles; Generating an individual cognitive entangled state of the local terrain based on the current observation data, wherein the individual cognitive entangled state comprises geometric virtual forms of at least two alternative terrains and corresponding confidence degrees, and the individual cognitive entangled state is used for representing quantum mixed state representation of each unmanned aerial vehicle on the local terrain; performing consensus evolution on each individual cognitive entangled state aiming at the same local topography to generate a group cognitive field, wherein the group cognitive field is used for representing a geometric distribution field of the local topography; each unmanned aerial vehicle evaluates the cognitive conflict level based on the current state of the group cognitive field to obtain an expected excitation value; based on the expected excitation values, an observation target and an observation path are determined for topographic mapping from the observation target and the observation path. Optionally, the generating an individual cognitive entanglement state for the local terrain based on the current observation data includes: performing coherent computation on the current observation data, and extracting topological features of the local topography; Inputting the topological feature into a preset geometric virtual form generating network, converting the topological feature into the geometric virtual form by a tensor network in the preset geometric virtual form generating network, and outputting the geometric virtual form and confidence, wherein the confidence is used for representing entanglement entropy of the geometric virtual form. Optionally, before the inputting the topological feature into the preset geometric virtual form generating network, the method further comprises: Generating an countermeasure network by utilizing a preset condition based on the current observation data and the corresponding fact observation condition vector, and generating a counterfact observation condition vector, wherein the counterfact observation condition vector comprises at least one of disturbance factors, geometric shielding scenes and sensor characte