Road Detection Using Semantic Segmentation and Historical Position Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current road detection methods, especially those using satellite images, face inefficiencies and accuracy issues due to obstructions like clouds and shadows, and manual data collection is time-consuming.
Innovation Solution
A road detection method and apparatus that combines image semantic segmentation using a pre-trained fully convolutional network and conditional random field models with historical position information from terminals to improve accuracy by weighting probabilities and filtering noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If satellite image processing is used for road detection, then automation is improved, but detection accuracy deteriorates due to obstructions like clouds and shadows
Solution Approach 1:
The patent combines satellite image processing with historical position data from multiple terminals to compensate for obstructions. When clouds or shadows obscure roads in satellite images, the system merges this with trajectory data from historical positions to maintain detection accuracy while preserving automation.
Solution Approach 2:
Historical position information acts as an intermediary to bridge the gap caused by satellite image obstructions. The system uses this intermediate data source to verify and supplement road detection when primary satellite imagery is compromised by clouds or shadows.
2Measurement precision
If manual data collection is used for road detection, then detection accuracy is maintained, but productivity deteriorates due to time-consuming processes
Solution Approach 1:
The system enables self-service by automatically collecting and processing historical position data from multiple terminals without requiring manual intervention. This automated approach maintains high detection accuracy through data fusion while dramatically improving productivity by eliminating time-consuming manual collection processes.
Solution Approach 2:
The system implements feedback mechanisms where historical position data continuously informs and refines road detection. This automated feedback loop maintains accuracy comparable to manual methods while achieving high productivity through systematic data collection and processing from multiple sources.
3Measurement precision
If historical position information is integrated with image segmentation, then detection accuracy is improved, but device complexity increases
Solution Approach 1:
The system achieves multi-functionality by using a unified processing framework that handles both satellite image segmentation and historical position data integration. This universal approach improves detection accuracy through combined data sources while managing complexity through a consolidated system architecture rather than separate specialized systems.
Data Source
AI summary
A road detection method and apparatus. A specific embodiment of the method includes: acquiring an image of a predetermined region; semantically segmenting the image to acquire a first probability that a region corresponding to each pixel in the image is a road region; acquiring a historical position information set of a target terminal; correcting, in response to historical position information existing in the historical position information set, the historical position information indicating a historical position located in the predetermined region, the first probability according to the historical position information to obtain a second probability; and determining a region corresponding to a pixel having the second probability greater than a preset threshold as a road region. Such an embodiment improves the road detection accuracy.