Traffic Guide Object Recognition Using Aspect Ratio Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic guide object recognition systems face challenges in accurately recognizing traffic guide objects, especially in anomalous crossings where roads are not orthogonal and multiple traffic guide objects are present, leading to reduced recognition accuracy.
Innovation Solution
A traffic guide object recognition device and method that uses an imager to capture surroundings and a recognizer to select the appropriate traffic guide object by analyzing forms, directions, and distances of traffic guide objects, including traffic signals with light emitters and hood portions, based on image processing and map information, to improve recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional traffic guide object recognition is used, then simple road scenarios can be handled, but recognition accuracy deteriorates in anomalous crossings with non-orthogonal roads or multiple lanes
Solution Approach 1:
The patent introduces a new dimension of analysis by examining the aspect ratios of image areas occupied by traffic guide objects. Instead of relying solely on positional information, the system derives directions from aspect ratios, adding a dimensional characteristic that enables accurate recognition in anomalous crossing scenarios where traditional methods fail.
Solution Approach 2:
The system changes the parameter used for direction derivation from simple positional coordinates to aspect ratio measurements. By transforming the recognition parameter from location-based to shape-based analysis, the system achieves improved adaptability to various road configurations while maintaining high recognition accuracy.
2Adaptability or versatility
If multiple traffic guide objects are present near the vehicle, then more comprehensive coverage is provided, but selecting the correct object to follow becomes more difficult
Solution Approach 1:
The system employs feedback mechanisms by continuously comparing the derived directions of multiple traffic guide objects with the vehicle's traveling direction. The recognizer selects the object whose direction best matches the vehicle's direction, using this feedback loop to accurately identify the target traffic guide object even when multiple objects are present in the vicinity.
Solution Approach 2:
The patent applies local quality by deriving directions individually for each traffic guide object based on its specific aspect ratio characteristics. Instead of treating all objects uniformly, the system analyzes each object's local geometric properties to determine its direction, enabling precise differentiation and selection among multiple objects.
Data Source
AI summary
A traffic guide object recognition device includes an imager configured to image surroundings of a vehicle and a recognizer configured to recognize surrounding circumstances of the vehicle. The recognizer selects a traffic guide object which the vehicle is to follow on the basis of forms of traffic guide objects in an image captured by the imager out of traffic guide objects which are provided near the vehicle.