Electronic Map POI Tagging via Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for collecting and processing point of interest (POI) data in electronic maps are inefficient, requiring professional teams and taking a long time due to the need for manual data collection, processing, and transfer, which limits the accuracy and speed of updating POIs.
Innovation Solution
A method and system that utilize a terminal to acquire images and determine if objects within those images are already tagged on an electronic map, allowing for quick recognition and tagging of new POIs by searching historical images associated with the capturing position, thereby improving data processing efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If professional teams manually collect and process POI data, then data accuracy is improved, but processing time and cost increase significantly
Solution Approach 1:
The patent replaces manual mechanical data collection processes with automated image recognition technology. Mobile terminals capture images of POIs, and server-side recognition systems automatically identify and extract POI information from these images, eliminating the need for professional teams to manually collect and process data while maintaining high accuracy
Solution Approach 2:
The patent uses image copying as an intermediary - instead of directly collecting structured POI data through manual methods, the system captures visual copies (images) of POIs using mobile terminals, then automatically converts these image copies into structured data through recognition algorithms, significantly reducing processing time
2Reliability
If professional teams manually process POI data, then data quality is improved, but labor cost and operational complexity increase
Solution Approach 1:
The patent enables the system to serve itself by automatically performing data collection, recognition, and processing tasks. Mobile terminals automatically capture images, and the server automatically recognizes and processes POI information without requiring professional teams, thereby maintaining data quality while reducing operational complexity
Solution Approach 2:
The patent creates a universal system where mobile terminals can serve multiple functions - acting as both data collection devices and image capture devices. The server-side system provides universal processing capabilities that handle various types of POI data through a single automated recognition pipeline, reducing the need for specialized manual processing procedures
3Measurement precision
If traditional manual methods are used for POI collection, then data verification is thorough, but processing speed decreases
Solution Approach 1:
The patent enables continuous automated processing of POI data through image recognition. Multiple mobile terminals can continuously capture images of POIs, and the server continuously processes these images without interruption, maintaining thorough verification through automated recognition algorithms while achieving high-speed continuous data acquisition
Solution Approach 2:
The patent performs preliminary image capture and transmission before full processing. Mobile terminals pre-capture images and transmit them to the server, which then performs recognition and verification in advance, enabling rapid POI acquisition by having data ready before it is actually needed for map updates
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
The present disclosure provides a method, a device and a system for processing image tagging information. The method includes the following. A captured image and a capturing position are acquired from a terminal (101; 201). Object recognition is applied to the captured image to determine a first object presented by the captured image (102; 202). A first target image presenting the first object is searched for among one or more historical images associated with the capturing position (103; 205). When the first target image is searched for, it is determined that the first object is duplicated with tagging information of an electronic map (104; 209). When the first target image is not searched for, a corresponding position of the first object on the electronic map is tagged (105; 210).