Sensor Data Evaluation for Filtering Phantom Objects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Driver assistance and automated driving systems face limitations due to high false-positive rates from sensors, often caused by phantom objects formed by reflections or multiple reflections, which can significantly impact system availability.
Innovation Solution
A method for evaluating sensor data that identifies and filters objects based on surface characteristics using a trainable classifier and database access, allowing for improved accuracy and reliability by distinguishing between real and phantom objects, and reducing false-positive rates without affecting correct-positive rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If object detection is performed using radar sensors, lidar sensors and cameras to achieve sufficient correct-positive rate, then the reliability of object recognition is improved, but the false-positive rate increases due to phantom objects formed by reflection or multiple reflections
Solution Approach 1:
The patent applies the principle of analyzing surface characteristics (analogous to color changes) by examining reflectivity properties of detected objects. The system determines surface characteristics of detected objects and compares them with expected characteristics from map data to identify phantom objects caused by reflections, thereby reducing false positives while maintaining reliable object recognition.
Solution Approach 2:
The patent introduces map data as an intermediary element to verify detected objects. By comparing detected object characteristics with pre-stored map information about surface properties of known objects and areas, the system can distinguish real objects from phantom reflections, reducing false positives without compromising recognition reliability.
2Reliability
If a high false-positive rate is accepted to achieve sufficient correct-positive rate in object detection, then the reliability of sensor systems is improved, but the availability of downstream functions is considerably limited
Solution Approach 1:
The patent performs preliminary verification of detected objects by determining their surface characteristics and comparing them with map data before passing results to downstream functions. This preliminary action filters out phantom objects early in the processing chain, maintaining high correct-positive rates while preventing false positives from degrading downstream function availability.
Solution Approach 2:
The system implements feedback by using map data to verify detected objects and provide correction information. The comparison between detected surface characteristics and expected characteristics from map data creates a feedback mechanism that reduces false positives, thereby maintaining both reliability and downstream function availability.
3Measurement precision
If surface characteristics of detected objects are determined and compared with map data to reduce phantom objects, then the false-positive rate decreases, but the device complexity increases due to additional processing steps
Solution Approach 1:
The patent applies multi-functionality by using map data for multiple purposes: navigation, location verification, and object characteristic verification. The same map data structure serves both positional reference and surface property reference, reducing the need for separate verification systems and minimizing additional complexity while effectively reducing false positives.
Data Source
AI summary
A method for evaluating sensor data. The sensor data are ascertained by scanning a surrounding area, using at least one sensor. On the basis of the sensor data, object detection is carried out for determining objects from the sensor data. Object filtering is carried out. Surface characteristics of at least one object are identified, and/or the surface characteristics of at least one object are ascertained with the aid of access to a database. A control unit is also described.