Vehicle Signal Light State Detection Using Temporal Classifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques fail to accurately identify the ON and OFF states of a vehicle's signal lights from images, especially when the cycle and duration of these states are indefinite, making it difficult for automated driving systems to predict vehicle motion effectively.
Innovation Solution
An apparatus and method using a processor to input time series images into a first classifier to detect vehicle regions, and then into a second classifier with a recursive structure or performing convolution operations in the temporal direction to calculate confidence scores for candidate states of the signal light, allowing accurate identification of ON and OFF states based on transition information and confidence scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional object detection techniques are used to detect signal light states, then the detection process is simple, but the identification accuracy of signal light ON/OFF states is insufficient
Solution Approach 1:
The detection system is divided into two specialized classifiers: a first classifier for detecting vehicle presence and a second classifier for detecting signal light states. This segmentation allows each classifier to focus on its specific task, improving overall detection accuracy while maintaining manageable complexity through functional division.
Solution Approach 2:
The second classifier introduces a temporal dimension by processing time-series images sequentially and incorporating previous detection results. This transforms the detection from a single-frame static analysis to a multi-frame dynamic analysis, significantly improving signal light state identification accuracy by considering temporal continuity and state transitions.
2Measurement precision
If single-frame image detection is used, then the processing speed is fast, but the accuracy of identifying signal light states with indefinite cycles is insufficient
Solution Approach 1:
The system performs preliminary detection using the first classifier to identify vehicle regions before applying the more computationally intensive second classifier. This preliminary filtering reduces the amount of data requiring detailed analysis, maintaining processing efficiency while enabling accurate multi-frame temporal analysis for signal light state identification.
Solution Approach 2:
The system incorporates feedback from previous detection results into the current frame analysis. The second classifier uses the detected signal light states from previous frames to inform current state determination, allowing the system to leverage temporal patterns and improve accuracy for signal lights with indefinite cycles while maintaining efficient processing through intelligent reuse of historical data.
Data Source
AI summary
An apparatus for identifying the state of a signal light includes a processor configured to input time series images into a first classifier to detect object regions each including a vehicle equipped with a signal light in the respective images, the first classifier having been trained to detect the vehicle; chronologically input characteristics obtained from pixel values of the object regions detected in the respective images into a second classifier to calculate confidence scores of possible candidate states of the signal light of the vehicle, the second classifier having a recursive structure or performing a convolution operation in a temporal direction; and identify the state of the signal light, based on the preceding state of the signal light, information indicating whether transitions between the candidate states of the signal light are allowed, and the confidence scores of the respective candidate states.