Vehicle Object Recognition Using Light Patterns at Night
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic vehicle following systems face challenges in accurately recognizing objects at night due to reduced recognition abilities compared to daytime conditions.
Innovation Solution
An object recognition apparatus mounted in a vehicle, equipped with sensors like cameras, radars, and sonars, detects objects and determines their type based on the arrangement and number of lights, allowing for accurate identification of two-wheeled and four-wheeled vehicles by analyzing light patterns even in low-light conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If object recognition relies on clear object outlines captured by camera, then recognition accuracy is maintained during daytime, but recognition ability deteriorates during nighttime due to reduced visibility
Solution Approach 1:
The patent introduces light distribution patterns as an intermediary feature for object recognition. Instead of relying directly on object outlines which become invisible at night, the system uses the pattern of lights emitted by objects (taillights, headlights, position lights) as a mediator that remains detectable in low-light conditions. The sensor detects these light patterns and the determination unit classifies objects based on their characteristic light distributions, enabling nighttime recognition without depending on ambient illumination or object outlines.
2Reliability
If the system uses multiple sensors to detect objects, then detection capability is improved, but the complexity of determining object types increases
Solution Approach 1:
The patent segments the object type determination process into distinct categories based on light distribution patterns. The determination unit divides vehicle types into specific groups (two-wheeled vehicles with vertical light arrays, four-wheeled vehicles with horizontal light arrays, motorcycles with specific light configurations) and assigns recognition rules to each segment. This segmentation simplifies the overall complexity by breaking down the complex task of object identification into manageable, rule-based classifications rather than requiring a single complex recognition algorithm.
Data Source
AI summary
An object recognition apparatus mounted in a vehicle including a sensor is provided. The apparatus includes a detection unit configured to detect an object present in a same lane as that of the vehicle by using information from the sensor, an acquisition unit configured to acquire information concerning lights which the object turns on by using the information from the sensor, and a determining unit configured to determine a type of the object based on the information concerning the lights. A condition in which the determination unit determines that the type of the object is a two-wheeled vehicle includes a case in which the object includes not less than two lights arrayed in a vertical direction with respect to the ground surface.