LiDAR Discrete Object Detection via Grid Slope Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying on-ground or near-ground discrete objects using LiDAR data face challenges such as inefficiency in processing large datasets and distinguishing objects from surrounding noise, especially in inaccessible areas and when extensive field calibration is infeasible.
Innovation Solution
A method involving the analysis of LiDAR height data to determine underlying trends and distinguish individual objects with specific characteristics by forming a regular grid with elevation and slope data, using a Ground Feature Transformation (GFT) function to identify discrete objects, and employing nonconventional classification schemes to detect both on-ground and above-ground objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If LiDAR data is used to survey large or inaccessible areas, then the coverage area is improved, but the ability to distinguish objects from surrounding noise deteriorates
Solution Approach 1:
The patent divides the LiDAR data processing into multiple stages: initial object candidate identification, filtering based on characteristic features (size, shape, height), and validation against surrounding terrain patterns. This segmentation allows the system to maintain high detection precision across large areas by processing data in manageable steps rather than attempting simultaneous analysis of the entire dataset.
Solution Approach 2:
The patent applies different analysis methods and criteria to different regions of the LiDAR data based on local terrain characteristics. By adapting the detection parameters and feature extraction methods to match local ground conditions, the system maintains high object detection precision even when surveying diverse and extensive areas with varying noise patterns.
2Reliability
If traditional LiDAR processing methods are used, then the detection capability is maintained, but the processing efficiency deteriorates
Solution Approach 1:
The patent performs preliminary processing of LiDAR data by pre-calculating and storing characteristic features such as elevation, slope, aspect, and curvature for each data point before actual object detection begins. This preliminary action creates a pre-processed dataset that can be quickly queried and analyzed, dramatically improving processing efficiency while maintaining reliable detection capability through the use of pre-computed geometric features.
Solution Approach 2:
The patent transforms the raw LiDAR point cloud data into derived parameters including elevation maps, slope angles, aspect directions, and curvature values. By changing from raw coordinate data to these meaningful geometric parameters, the system enables faster object detection while preserving the ability to accurately identify objects based on their characteristic geometric properties.
3Measurement precision
If extensive field calibration is performed, then the detection accuracy is improved, but the time and resource requirements worsen
Solution Approach 1:
The patent implements self-calibration through automated algorithms that use the LiDAR data itself to establish detection thresholds and parameters. The system automatically identifies terrain patterns, calculates statistical distributions of ground features, and sets detection criteria without requiring manual field calibration, thereby achieving high detection accuracy while eliminating the time-consuming field calibration process.
Solution Approach 2:
The patent performs all necessary calibration and parameter optimization during the data processing phase rather than requiring separate field calibration activities. By pre-establishing detection thresholds and validating algorithms against known terrain patterns in the LiDAR data itself, the system achieves accurate object detection without the need for time-consuming field calibration trips to inaccessible areas.
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
This approach enables efficient detection of discrete objects by accurately distinguishing them from the surrounding terrain, even in large or inaccessible areas, improving the identification of on-ground and near-ground features with high precision.
Implementation Method 1
The LiDAR unit transmits and detects laser pulses which are reflected off objects on the ground or in the air
Implementation Method 2
The LiDAR unit measures the time, direction and strength of each reflected laser pulse
Data Source
AI summary
A method for detection of on-ground or near-ground discrete objects having characteristic features from terrain height data characterizing a search area comprising the steps of: processing the terrain height data to form a regular grid containing elevation and slope data at each grid point; and analyzing said gridded terrain height data to identify said on-ground or near-ground discrete objects within said search area.