Autonomous Vehicle Delta Vision for Low-Power Obstacle Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional control systems for autonomous and semi-autonomous vehicles face challenges in achieving near-perfect safety due to the complexity of handling unforeseen obstacles and high-speed interactions, which requires extensive sensor arrays and significant processing power, leading to cost and energy efficiency issues.
Innovation Solution
The implementation of delta imaging and artificial neural networks to identify objects and determine their characteristics within autonomous and semi-autonomous vehicles, allowing for reduced sensor reliance and processing power by focusing on delta information, enabling efficient object detection and control operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If numerous sensors and sensor types are used to ensure near-perfect safety, then safety reliability is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts and focuses only on the essential information needed for safety - delta information representing changes in the environment. By using delta imaging to capture only changes between frames rather than processing complete images from multiple sensors, the system identifies and processes only the critical data needed for obstacle detection, thereby reducing sensor requirements while maintaining safety
Solution Approach 2:
The patent segments the visual information processing by dividing complete images into delta information representing only changes. This segmentation allows the system to process only the relevant portions of visual data (changes in pixel values between frames) rather than processing entire images from multiple sensors, reducing computational complexity and sensor requirements
2Measurement precision
If numerous sensors are equipped to autonomous vehicles, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The patent extracts only the essential delta information from complete images - specifically the changes in pixel values between frames. By processing only this extracted delta information rather than complete high-resolution images from multiple sensors, the system maintains object detection precision while significantly reducing the energy required for image processing
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of visual information (delta changes) rather than complete images. This partial processing approach maintains sufficient measurement precision for safety-critical object detection while reducing energy consumption by avoiding processing of redundant unchanged portions of images
3Reliability
If massive computer processing power is used to interpret sensor data in real-time, then reliability is improved, but use of energy and device complexity increase
Solution Approach 1:
The patent extracts only delta information representing changes from complete images before processing. By feeding only this extracted delta information into neural networks and processing systems, the computational load is dramatically reduced while maintaining real-time processing capability and reliability for safety-critical decisions
Solution Approach 2:
The patent performs preliminary processing by generating delta images and extracting change information before the main processing stage. This preliminary extraction of essential information reduces the data volume that requires massive processing power, enabling real-time reliable control with reduced energy consumption
Data Source
AI summary
Methods and systems are disclosed for an improved control system in autonomous and semi-autonomous vehicles. More specifically, the methods and systems relate to powering control systems of autonomous and semi-autonomous vehicles through the use of computer vision based on delta images (i.e., delta-vision).