Octree Data Structure for Vehicle Sensor Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data organization schemes for vehicle sensors are inefficient in representing and classifying a wide variety of inputs and data types, limiting their usefulness for autonomous mission execution and automation in vehicles.
Innovation Solution
The proposed solution involves organizing raw sensor data using an octree structure with super nodes that link to classified data objects stored in a database, enabling real-time or near real-time classification and efficient storage and retrieval of relevant data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If current data organization schemes are used for vehicle sensors, then data storage is simple, but data representation and classification efficiency is poor
Solution Approach 1:
The patent segments sensor data into hierarchical groups and subgroups (e.g., terrain, vegetation, water bodies, man-made structures) with further classification into detailed categories. This segmentation enables efficient classification and representation of diverse sensor data types while maintaining organized structure through the octree data structure that divides spatial data into manageable nodes.
Solution Approach 2:
The patent introduces a hierarchical dimensional structure using octree organization where data is arranged in multiple levels from root nodes to leaf nodes, adding temporal and spatial dimensions to data organization. This multi-dimensional approach enables efficient querying and classification by allowing data to be accessed and classified from different organizational perspectives simultaneously.
2Adaptability or versatility
If a unified data organization scheme is created for diverse sensor inputs, then classification capability improves, but system complexity increases
Solution Approach 1:
The patent creates a universal data organization scheme that can handle multiple sensor types (LIDAR, optical, infrared, radar) and diverse data formats through a single octree-based hierarchical structure. The standardized classification framework with predefined groups and subgroups provides multi-functional capability to organize various terrain, vegetation, water body, and man-made structure data uniformly, reducing the need for separate organization systems for different sensor types.
Solution Approach 2:
The patent employs parameter-based classification where sensor data is organized according to multiple parameters including terrain characteristics, vegetation density, water body properties, and man-made structure types. By changing and combining different classification parameters at various hierarchical levels, the system achieves versatile adaptability to handle diverse data types while maintaining a consistent organizational framework.
3Extent of automation
If detailed classification of sensor data is implemented, then autonomous mission execution capability improves, but data processing time increases
Solution Approach 1:
The patent implements preliminary classification by pre-defining groups and subgroups for different terrain types, vegetation, water bodies, and man-made structures before actual sensor data processing. The octree data structure is pre-configured with classification categories, enabling rapid data organization and interpretation during autonomous missions without requiring complex real-time classification decisions, thus reducing processing time while maintaining detailed classification capability.
Solution Approach 2:
The patent uses template-based classification where standardized data models and classification categories are created as reusable templates for different terrain and object types. Once a classification scheme is established for a particular terrain type or object category, it can be copied and applied to similar data, significantly reducing processing time for repetitive classification tasks while maintaining consistent and detailed classification across diverse datasets.
Data Source
AI summary
An apparatus is described comprising at least one processor; and memory storing instructions that, when executed by the at least one processor, cause the apparatus to: organize items of raw data received from at least one sensor of a vehicle as a first data structure, organize classified data objects as a second data structure, and link at least one item of the first data structure to at least one object of the second data structure.