Antenna Downtilt Angle Measurement Using UAV and Deep Semantic Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for measuring antenna downtilt angles in mobile communication networks are labor-intensive, costly, and not suitable for large-scale measurements, posing safety hazards and inefficiencies due to manual intervention and instrument installation challenges.
Innovation Solution
A method utilizing a multi-scale deep semantic segmentation network, combined with an unmanned aerial vehicle, to collect and analyze images for target detection and semantic segmentation, enabling accurate calculation of antenna downtilt angles through image processing and geometric calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual measurement methods are used to measure antenna downtilt angles, then measurement accuracy can be maintained, but safety hazards and workload increase significantly
Solution Approach 1:
The patent replaces manual mechanical measurement methods with an automated image processing system. A terminal device captures images of the antenna, and a server performs automated analysis to determine the downtilt angle, eliminating the need for personnel to physically access high-altitude locations while maintaining measurement accuracy through computational geometry calculations.
Solution Approach 2:
The patent creates a visual copy of the antenna through image capture. Instead of direct physical measurement, the system captures an image of the antenna and its mounting structure, then analyzes the image data to extract geometric information and calculate the downtilt angle, replacing physical interaction with optical documentation.
2Productivity
If angle sensors are installed on antennas to measure downtilt angles, then regular measurement can be performed, but installation time and cost increase
Solution Approach 1:
The patent extracts the measurement function from physical instruments installed on the antenna. Instead of attaching angle sensors to the antenna structure, the system extracts geometric information from images of the antenna and its mounting bracket, performing calculations on a remote server to determine the downtilt angle without any physical modification to the antenna.
Solution Approach 2:
The patent creates a universal measurement system that can measure downtilt angles of different antenna types without requiring specific instrumentation for each antenna model. The image-based approach works across various antenna configurations, eliminating the need for model-specific sensor installations while enabling regular measurement of multiple antennas.
3Measurement precision
If traditional measurement methods are used for large numbers of antennas, then individual measurement accuracy is maintained, but resource consumption becomes unsustainable
Solution Approach 1:
The patent merges multiple measurement operations into a single automated workflow. The system captures images of multiple antennas and processes them through a unified image analysis pipeline on the server, calculating downtilt angles for numerous antennas using the same computational geometry algorithms, thereby maintaining individual measurement accuracy while dramatically reducing the manpower and material resources required compared to traditional methods.
Data Source
AI summary
A method for measuring an antenna downtilt angle based on a multi-scale deep semantic segmentation network is disclosed, including: collecting base station antenna data by using an unmanned aerial vehicle, and labeling an acquired antenna image with a labeling tool to make a data set; calling the data set for training and debugging a model; recognizing and detecting a target antenna, performing semantic segmentation on an output image, finally obtaining a target image finally segmented, and calculating a downtilt angle of the target image. The method is highly applicable, cost-effective, and safe.