Lidar Data Processing
The method uses moving averages to classify LIDAR sensor data, distinguishing between active and inactive blanks to accurately detect vehicles with low-reflectivity surfaces, enhancing real-time processing and reducing misclassification.
Patent Information
- Authority / Receiving Office
- GB · GB
- Patent Type
- Patents
- Current Assignee / Owner
- RED FOX I D LTD
- Filing Date
- 2023-11-27
- Publication Date
- 2026-06-24
AI Technical Summary
LIDAR sensors struggle with distinguishing between low-reflectivity surfaces and sensor faults, leading to incorrect classification of vehicles, particularly with articulated vehicles, where the cab is detected but the trailer is misclassified as a separate vehicle due to blank readings.
A method using moving averages to classify blank data points as 'active' or 'inactive' based on the frequency of blank and non-blank readings, allowing for accurate detection of vehicles by interpreting 'active blanks' as indicative of low-reflectivity surfaces, and adjusting actions like camera triggering accordingly.
Enables accurate vehicle detection and timely follow-up actions, such as photographing license plates, by correctly identifying vehicles with low-reflectivity surfaces, reducing misclassification and improving real-time data processing efficiency.
Smart Images
Figure 00000001_0000 
Figure 00000002_0000