Lidar Object Motion Analytics via Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for analyzing lidar sensor data struggle with accurately processing and analyzing dynamic targets with objects in motion, particularly in providing comprehensive object motion analytics.
Innovation Solution
A system comprising a network-connected computer and lidar sensors that processes lidar point cloud data to generate object lists with classification, motion, position, and size information, using connected components-based graph processing, machine learning models, and Kalman filters to track and classify moving objects, and synchronize data across sensors for comprehensive object motion analytics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional static target analysis techniques are used, then the analysis process is simple, but the accuracy of analyzing dynamic targets with objects in motion deteriorates
Solution Approach 1:
The patent segments the dynamic target analysis into distinct phases: generating point cloud data from multiple lidar sensors, clustering points into objects, tracking object motion across frames, and classifying objects based on motion patterns. This segmentation allows each sub-process to be optimized independently, improving overall accuracy while managing complexity through modular processing stages.
Solution Approach 2:
The system performs preliminary actions by pre-processing lidar data into point clouds, pre-clustering points into potential objects, and establishing tracking frameworks before actual motion analysis. This preliminary structuring of data enables more accurate motion detection and classification while reducing the computational complexity during real-time analysis.
2Reliability
If comprehensive object motion analytics are generated, then the surveillance capability is enhanced, but the processing time and computational load increase
Solution Approach 1:
The patent applies partial action by focusing computational resources on tracking and analyzing only moving objects rather than processing all detected objects equally. The system identifies objects with motion characteristics and applies comprehensive analytics only to those, while static objects receive minimal processing. This selective approach maintains high surveillance capability for dynamic targets while reducing overall processing time and computational load.
3Area of stationary object
If multiple lidar sensors are used to improve coverage, then the geolocation coverage is enhanced, but the data synchronization complexity increases
Solution Approach 1:
The patent merges data from multiple lidar sensors by integrating point clouds from different sensor locations into a unified coordinate system. The system combines detection results from multiple sensors, consolidates object tracks across sensor fields of view, and produces unified motion analytics. This merging approach enhances geolocation coverage while managing synchronization complexity through integrated data fusion rather than independent sensor processing.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively generates accurate object motion analytics, enabling efficient tracking and classification of moving objects, and provides predictive analytics such as people counting and queue analysis, enhancing surveillance capabilities in various environments like airports.
Implementation Method 1
Light detection and ranging (Lidar) is a method for measuring distances by illuminating a target with laser light and measuring the reflection with a sensor
Implementation Method 2
Light detection and ranging (Lidar) is a method for measuring distances by illuminating a target with laser light and measuring the reflection with a sensor
Data Source
AI summary
A system has a collection of lidar sensors to generate lidar point cloud data for a defined geolocation. A computer is connected to the collection of lidar sensors via a network. The computer includes a processor and a memory storing instructions executed by the processor to process the lidar point cloud data to produce an object list for each moving object identified in the lidar point cloud data. The object list includes object classification information, object motion information, object position data, and object size data. Object motion analyses are performed on each object list to generate object motion analytics for the defined geolocation.