Sensor-Readable Tags for Accurate Map Landmark Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image sensors acquire raster images without attribute information of landmarks, requiring post-processing to overlay geographical label layers, which can result in incomplete, outdated, or inaccurate landmark names due to methods like digitizer-based or crowdsourcing approaches.
Innovation Solution
The auto-generation of map landmarks using sensor-readable tags (SRTs), where physical tags encoded with landmark information are captured by sensors, decoded, and used to generate a geographical label layer for real-time annotation of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If digitizer-based or crowdsourcing approaches are used to build geographical label layers, then landmark information can be obtained, but the information may be incomplete, outdated, or inaccurate
Solution Approach 1:
The system enables landmarks to self-announce their own information through SRTs that encode landmark identifiers and attributes. When sensors capture images of these tags, the landmark information is automatically extracted and added to the geographical label layer without requiring human digitizers or crowdsourcing, thereby ensuring complete and accurate information.
2Productivity
If traditional methods are used to generate landmark information, then the process can be completed, but it is time-consuming and results in outdated data
Solution Approach 1:
The patent replaces manual mechanical processes of digitizers and crowdsourcing with an automated sensor-based system. Sensors capture images of SRTs, which are then processed through image processing and decoding algorithms to automatically extract landmark information, dramatically increasing productivity and eliminating time delays associated with human-based data collection methods.
Solution Approach 2:
Landmark information is pre-encoded into SRTs that are placed at landmark locations before imaging. This preliminary encoding allows sensors to directly capture ready-to-use landmark data during routine imaging operations, eliminating the need for separate data collection processes and enabling real-time updates without time loss.
3Productivity
If raster images are acquired without attribute information, then image capture is simple and fast, but the images lack landmark annotations for intelligent applications
Solution Approach 1:
The system merges image capture with landmark information acquisition by integrating SRT detection and decoding into the imaging workflow. Sensors simultaneously capture both the visual scene and the encoded landmark tags within the same image, allowing rapid image acquisition while automatically extracting and attaching landmark attribute information to the corresponding image regions.
Data Source
AI summary
An image of a geographical area is generated, the geographical area including at least one sensor-readable tag (SRT). The image of the geographical area includes images of the at least one SRT. Each SRT is associated with a landmark and encoded with label information of the landmark. Images of the at least one SRT are extracted from the image of the geographical area. For each of the at least one SRT, the label information of the landmark is decoded based on the extracted image of the SRT. A geographical label layer is generated including the label information of each landmark. A superimposed image is generated by superimposing the geographical label layer on the image of the geographical area.