Pedestrian Detection Using Adaptive Identification Dictionaries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional pedestrian detection systems face challenges in improving accuracy due to environmental factors, as loosening detection criteria to account for changes leads to incorrect detections, while tightening criteria reduces detectivity, creating a trade-off between detectivity and correctness.
Innovation Solution
A pedestrian detection device and method that utilize a vehicular camera, sonar, and radar to analyze images by storing multiple identification dictionaries representing pedestrian features, selecting the most suitable dictionary based on environmental conditions, and learning from detection results to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the determination criterion to determine whether an image includes the feature of the pedestrian is loosened to detect pedestrians under varying environmental conditions, then the pedestrian detectivity is improved, but the percentage of correct detections is reduced due to incorrect detections
Solution Approach 1:
The patent segments the pedestrian detection problem by dividing the image into multiple candidate regions and evaluating each region independently using identification dictionaries. This allows the system to apply different evaluation criteria to different regions, improving overall detection accuracy while maintaining high detectivity across varying environmental conditions.
Solution Approach 2:
The patent changes the parameters used for pedestrian identification by employing multiple identification dictionaries that represent different pedestrian features and characteristics. The system selects and applies appropriate dictionaries based on the specific image conditions, allowing flexible adaptation to environmental variations without sacrificing detection precision.
2Measurement precision
If the determination criterion to determine whether an image includes the feature of the pedestrian is made strict to avoid incorrect detections, then the percentage of correct detections is improved, but the pedestrian detectivity is reduced due to missed detections
Solution Approach 1:
The patent creates a universal detection system that uses multiple identification dictionaries to handle various pedestrian types and environmental conditions. This multi-functional approach allows the system to maintain strict evaluation criteria while still detecting diverse pedestrian forms, thereby improving both detection precision and detectivity simultaneously.
Solution Approach 2:
The patent introduces dynamic selection of identification dictionaries based on image characteristics and environmental conditions. The system adaptively adjusts which dictionaries are applied to different regions, allowing flexible modulation of detection strictness to optimize both precision and detectivity for each specific situation.
Data Source
AI summary
A pedestrian detection device on a vehicle having a vehicular camera and a sonar or a radar includes: an identification dictionary storage unit that stores multiple identification dictionaries of shape features of a pedestrian; a detection result accumulation unit that searches the captured image for one feature to detect the pedestrian, and accumulates a detection result for each identification dictionary; a pedestrian position accumulation unit that detects the pedestrian using the sonar or the radar, and accumulates a pedestrian position; an identification dictionary selection unit that selects a detection identification dictionary having the detection result consistent with the pedestrian position; and a pedestrian detection unit that searches the captured image for one feature described in the detection identification dictionary, and detects the pedestrian.