Fused Sensor Attribute Detection for Autonomous Vehicle Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately predicting object behavior due to limitations in sensor data integration, particularly in distinguishing between different objects using cameras and LiDAR/Radar systems, which affects anticipatory planning and control.
Innovation Solution
A system that processes observation probability distributions from multiple sensors to determine the probability of object attributes, using a tree structure and Bayesian filters to fuse data and adjust vehicle operations based on predicted trajectories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple sensors (camera, LiDAR, radar) are used for object detection, then the ability to distinguish between different object types improves, but the complexity of fusing sensor data and determining object attributes increases
Solution Approach 1:
The patent segments the object detection process into distinct probabilistic components: object type probability distribution, attribute detection probability, and attribute presence probability. Each sensor modality (camera, LiDAR, radar) contributes to specific segments of this probabilistic framework, allowing the system to manage complexity by treating each detection aspect separately while integrating results through probability fusion.
Solution Approach 2:
The patent introduces probability distribution functions as intermediary representations between raw sensor data and final object characterization. Instead of directly fusing heterogeneous sensor data, the system transforms sensor observations into probability distributions that can be systematically combined, serving as a mediator that simplifies the integration of multi-sensor information.
2Measurement precision
If probabilistic fusion of multiple sensors is implemented, then the accuracy of attribute detection improves, but the computational processing requirements increase
Solution Approach 1:
The patent performs preliminary actions by pre-defining the probabilistic framework and detection models before actual sensor data arrives. The system establishes probability distribution functions and detection algorithms in advance, allowing real-time operation to focus on updating these pre-established models with incoming sensor data rather than building the entire detection pipeline from scratch during critical decision moments.
3Reliability
If the system determines both object type and attributes with high confidence, then the safety of autonomous driving improves, but the time required for processing increases
Solution Approach 1:
The patent applies partial action by determining object type and attributes independently through separate probabilistic pathways rather than requiring complete analysis of all sensor data for both aspects simultaneously. The system can achieve sufficient confidence in object type detection without fully resolving all attribute uncertainties, allowing timely decision-making while maintaining adequate safety margins.
Data Source
AI summary
A system receives an observation probability distribution function associated with a target object that was detected by sensors of an autonomous vehicle. The system identifies a target attribute of the target object, and detects a target attribute value associated with the target object. The system determines a first probability distribution function representing a probability of the autonomous vehicle detecting an object having an object label, determines a second probability distribution function defining a probability of the autonomous vehicle detecting the target attribute, determines a third probability distribution function defining a probability of the target attribute being present for the target object based on the target attribute value, and determines an attribute probability distribution function defining a probability that the target attribute is actually present for the target object. The system executes vehicle control instructions that cause the autonomous vehicle to adjust driving operations based on the attribute probability distribution function.