Vehicle Driving Control Using Object Passability Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control systems fail to accurately determine the drivability of objects in the front field, leading to potential unsafe maneuvers in emergency situations.
Innovation Solution
A method for controlling vehicle driving operations by sensing the front field using sensors and cameras, identifying relevant objects, and classifying them based on passability using a control unit and a specialized computing unit to determine the degree of passability, allowing for safe maneuvering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing vehicle control systems use basic sensor data for object detection, then the system complexity is kept low, but the measurement precision and reliability of determining object passability deteriorates
Solution Approach 1:
The system segments object detection into multiple processing stages: initial sensor detection, optical identification using cameras, point cloud generation for 3D positioning, and classification by a specialized computing unit. Each stage handles specific aspects of object analysis, improving overall precision while distributing computational complexity across modular components rather than requiring one monolithic complex system.
Solution Approach 2:
The patent introduces point clouds as an intermediary data structure between sensor detection and object classification. Point clouds transform raw sensor data into a standardized 3D representation that facilitates accurate object positioning and passability assessment, serving as a mediator that enhances measurement precision without directly increasing the complexity of the core control system.
2Reliability
If the vehicle uses comprehensive sensor data and environmental images for object identification, then the reliability of driving control improves, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary actions by continuously capturing environmental images and generating point clouds during normal driving operations, even before emergency situations arise. This pre-processing of sensor data ensures that when objects need rapid classification for safety decisions, the foundational data structures are already prepared, reducing critical processing time while maintaining high reliability.
Solution Approach 2:
The computing unit is specially configured with model parameters that enable it to autonomously classify objects based on pre-trained criteria. The system serves itself by having the computing unit independently evaluate point cloud data against stored passability models, eliminating the need for complex real-time analysis by the main control system and thereby reducing processing time while maintaining reliable decision-making.
3Reliability
If the system classifies objects as passable or non-passable with high accuracy, then the safety of vehicle maneuvers improves, but the device complexity for object classification increases
Solution Approach 1:
The patent extracts the complex classification function into a separate, specially configured computing unit that operates independently from the main vehicle control system. This computing unit contains dedicated model parameters for object passability assessment, isolating the complexity of high-accuracy classification from the core control architecture. The main system receives simplified passability decisions, maintaining maneuver safety while reducing overall system complexity.
Data Source
AI summary
The disclosure relates to a method for controlling a driving operation of a vehicle. According to the method, sensor data of a monitoring area in front of the vehicle is captured and environmental images of an environment of the vehicle are captured. Based on the sensor data and the environmental images, objects are optically identified and classified into known objects and unknown objects. A degree of passability is associated with the known objects. The unknown objects are further processed by input of their pictures into a computing unit configured with the degree of passability of objects. The unknown objects are identified and a degree of passability is associated with them. This degree of passability for the unknown objects is provided to a control unit and the driving operation of the vehicle is controlled by it by using the degree of passability of the objects and the unknown objects.