CN-122002129-A - Cloud AI-based full-automatic travel pat unmanned aerial vehicle system and method

CN122002129ACN 122002129 ACN122002129 ACN 122002129ACN-122002129-A

Abstract

The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to a cloud AI-based full-automatic travel pat unmanned aerial vehicle system and method. The system comprises an unmanned aerial vehicle end, an intelligent terminal APP, a cloud AI intelligent body and a built-in shooting master library, wherein the unmanned aerial vehicle end is carried by a user and does not comprise an AI chip and a cellular communication module and is used for establishing WiFi hot spots and executing photo and/or video shooting, the intelligent terminal APP is used for collecting user voice or text instructions, connecting the unmanned aerial vehicle WiFi and performing data transfer, and the cloud AI intelligent body is used for analyzing the user instructions, calling the built-in shooting master library to generate a shooting scheme and realizing automatic picture repair and video editing. According to the intelligent processing method, the intelligent processing function is moved up to the cloud end, so that the hardware cost of the unmanned aerial vehicle is greatly reduced, and the functions of voice control, automatic shooting, intelligent picture repairing, video editing, mirror transporting template, photo style template, multi-person group photo, image quality enhancement, one-key film forming, intelligent navigation returning, user preference learning and the like are realized, so that a common user can obtain high-quality travel photo works without professional learning, and the intelligent processing method has the advantages of being simple in operation, low in cost, good in shooting effect and the like.

Inventors

YANG JIANPING
YANG YANG

Assignees

杨洋

Dates

Publication Date: 20260508
Application Date: 20260325

Claims (12)

1. Full-automatic travel of AI based on high in clouds unmanned aerial vehicle system, its characterized in that includes: The unmanned aerial vehicle terminal is carried by a user, does not comprise a special chip for executing AI calculation and does not comprise a cellular communication module for accessing a public network, and is used for establishing a local area network wireless hotspot and executing a flight shooting task, wherein the flight shooting task comprises shooting pictures and/or videos; the intelligent terminal APP is installed on intelligent terminal equipment of a user, is used for collecting voice instructions and/or text instructions of the user, is connected with a local area network wireless hotspot established by the unmanned aerial vehicle end, and performs data transfer between the unmanned aerial vehicle end and the cloud AI intelligent agent; The cloud AI intelligent agent is deployed on the remote server and is used for receiving a user input instruction forwarded by the intelligent terminal APP, calling a built-in shooting master library to generate a shooting scheme containing flight tracks, shooting parameters and user guidance information, converting the scheme into an unmanned aerial vehicle executable instruction, and returning the unmanned aerial vehicle executable instruction to the unmanned aerial vehicle through the intelligent terminal APP.
2. The system of claim 1, wherein the drone side further comprises a microphone for capturing ambient sounds or as a back-up input for voice commands.
3. The system of claim 1 or 2, wherein the unmanned aerial vehicle further comprises an audio playing module, configured to play the user guidance information generated by the cloud AI agent.
4. The system of claim 1, wherein the intelligent terminal APP comprises a manual control module for generating control instructions in response to manual operation of a user on an APP interface and transmitting the control instructions to the drone side through the local area network wireless hotspot.
5. The system of claim 1, wherein the cloud AI agent comprises a shooting scheme generation module, wherein a shooting master library is built in the shooting scheme generation module, and the shooting master library comprises a composition rule library, a mirror track library, a light and shadow skill library and a gesture guidance library, and is used for calling corresponding rules from the shooting master library to generate a shooting scheme according to the voice command and/or the environmental information.
6. The system of claim 1, wherein the cloud AI agent comprises a slicing processing module, and is configured to perform automatic optimization processing on original material photographed by the unmanned aerial vehicle end, and generate an optimized slice, where the original material includes a photo and/or a video.
7. The system of claim 6, wherein the composition processing module supports a user to make modification requirements for a given composition via natural language input and to optimize the composition based on the modification requirements.
8. The system of claim 7, wherein the slicing process module employs a hierarchical optimization strategy that, for a first received modification request, preferentially employs an optimization mode of non-geometric deformation, including light and shadow adjustment, color optimization, and contrast enhancement, and when a second modification request is received, employs a geometric deformation optimization mode, including stretching legs, thinning faces, removing passers-by, replacing backgrounds, and supports multiple iterations, wherein the user can continuously propose multiple modification requests, and the system responds successively.
9. The system of claim 1, wherein the cloud AI agent comprises a user preference learning module configured to record historical operating information of a user and adjust according to the historical operating information during a subsequent imaging scenario generation or sheeting process.
10. The system of claim 1, wherein the cloud AI agent comprises a continuous shooting optimization module for responding to a shooting instruction of a user, controlling the unmanned aerial vehicle to collect a plurality of images or a video segment, and screening out the images or the video segment from the images or the video segment to be pushed to the user.
11. The system of claim 1, wherein the intelligent terminal APP further comprises a template selection module for displaying a plurality of mirror templates to a user, each template corresponding to a set of preset flight trajectories and shooting parameters, and responsive to a user selection instruction for a target template, the selection instruction is sent to a cloud AI agent, and the cloud AI agent invokes the corresponding mirror parameters in a photographic master knowledge base to generate flight control instructions.
12. A fully automatic beat method based on the system of any of claims 1-11, comprising the steps of: The unmanned aerial vehicle end establishes a local area network wireless hotspot; The intelligent terminal APP is connected with the local area network wireless hotspot; The user speaks a voice instruction through a microphone of the intelligent terminal APP or inputs a text instruction through a keyboard, and the intelligent terminal APP collects the user input instruction; The intelligent terminal APP forwards the user input instruction to a cloud AI intelligent agent through a wireless public network communication network; The cloud AI intelligent agent analyzes the user input instruction and calls a built-in shooting master library to generate a shooting scheme; the cloud AI intelligent agent converts the shooting scheme into an executable instruction of the unmanned aerial vehicle and forwards the executable instruction to the unmanned aerial vehicle through the intelligent terminal APP; the unmanned aerial vehicle end executes flight and shooting tasks according to the executable instructions, and photographs and/or videos are shot; the unmanned aerial vehicle side uploads the photographed original material to a cloud AI intelligent body through an intelligent terminal APP; the cloud AI intelligent agent optimizes the original materials to generate a sheet and pushes the sheet to the intelligent terminal APP, wherein the sheet comprises optimized photos and/or videos.

Description

Cloud AI-based full-automatic travel pat unmanned aerial vehicle system and method The invention belongs to the technical field of unmanned aerial vehicles, and particularly relates to an unmanned aerial vehicle system and method for intelligent shooting and intelligent picture repairing through cloud AI. Background The existing unmanned aerial vehicle has the following problems: 1. The price is high, and each unmanned aerial vehicle needs to be internally provided with an AI chip and a 4G/5G module, so that the cost is high. 2. The operation is complex, the flight skill and photographic knowledge are required to be specially learned, and the ordinary people are difficult to get on hand quickly. 3. The shooting effect depends on the personal level, and even if the person flies, the person does not know the professional shooting methods such as composition, fortune mirror and the like, and the quality of the shot photos and videos is common. 4. When one person goes out for travel, no one person helps to operate the unmanned aerial vehicle to take a picture or take a video. 5. The post-processing is troublesome, and the self-repairing and video editing are needed after the shooting is finished, so that the process is tedious. Disclosure of Invention The technical problem to be solved by the invention is that the existing unmanned aerial vehicle is high in cost (due to the built-in AI chip), complex in operation (professional learning is needed), incapable of being used by a single person (no help is taken), and complex in post-processing (self-repairing is needed). According to the invention, through moving the AI up to the cloud, carrying by a user, supporting voice/text direct control, internally arranging a shooting master library and layering optimization map repair, the technical problems are solved, and the 'cheap, simple, professional and single person available' travel pat unmanned aerial vehicle is realized. In order to achieve the above purpose, the present invention provides the following technical solutions: full-automatic travel of AI based on high in clouds unmanned aerial vehicle system includes: The unmanned aerial vehicle terminal is carried by a user, does not comprise a special chip for executing AI calculation and does not comprise a cellular communication module for accessing a public network, and is used for establishing a local area network wireless hotspot to execute a flight shooting task, wherein the flight shooting task comprises shooting pictures and/or videos. The unmanned aerial vehicle end comprises hardware such as a camera, a microphone, an audio playing module, a WiFi module, a battery, a motor and the like. The intelligent terminal APP is installed on intelligent terminal equipment of a user and used for collecting voice instructions and/or text instructions of the user, connecting the wireless hot spot of the local area network established by the unmanned aerial vehicle end and carrying out data transfer between the unmanned aerial vehicle end and the cloud AI intelligent agent. The APP supports a manual control function, and a user can control the unmanned aerial vehicle in real time through screen operation. The cloud AI intelligent agent is deployed on the remote server and is used for receiving a user input instruction forwarded by the intelligent terminal APP, calling a built-in shooting master library to generate a shooting scheme containing flight tracks, shooting parameters and user guidance information, converting the scheme into an unmanned aerial vehicle executable instruction, and returning the unmanned aerial vehicle executable instruction to the unmanned aerial vehicle through the intelligent terminal APP. Further, the cloud AI agent further includes: The shooting scheme generation module is internally provided with a shooting master library, and the shooting master library comprises a composition rule library, a fortune mirror track library, a light shadow skill library and a gesture setting guide library, and is used for calling corresponding rules from the shooting master library to generate a shooting scheme according to the voice command and/or the environmental information. The film forming processing module is used for automatically optimizing the original materials shot by the unmanned aerial vehicle end to generate optimized films, and the original materials comprise photos and/or videos. The module supports the user to make modification requirements for the designated film through natural language input, such as "lighten the photo 2", "cut the video 1 to 10 seconds", "remove the superfluous characters in the photo 1", and perform optimization processing according to the requirements. The slicing processing module adopts a layered optimization strategy, wherein for the first received modification requirement, optimization modes of non-geometric deformation, including light and shadow adjustment, color optimization and contrast enhancement, and when th