Scalable Video Compression for Autonomous Driving
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional video compression engines integrated in system on chip (SoC) are difficult to scale to meet the demands of autonomous driving vehicles (ADVs), which typically have multiple camera sensors, requiring a scalable, energy-efficient video compression solution.
Innovation Solution
A scalable video gathering and compression system is proposed, comprising a data gathering stage with interconnected data gathering nodes that combine image streams from multiple cameras and a data processing stage with nodes that use dedicated processing engines for efficient video compression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional video compression engines integrated in SoC are used, then the system can perform video compression, but the system cannot scale to meet the demands of autonomous driving vehicles with multiple camera sensors
Solution Approach 1:
The patent divides the video compression system into multiple independent processing nodes, each capable of handling compression tasks for specific camera streams. This segmentation allows the system to scale by adding or removing nodes based on the number of cameras, resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The system employs dynamic resource allocation where processing nodes can be activated or deactivated based on the actual number of camera streams being processed. This dynamic adaptability enables the system to scale flexibly without increasing permanent hardware complexity.
2Productivity
If CPU-based solutions are used for video compression, then the system can process video data, but the processing efficiency is insufficient and energy consumption is high
Solution Approach 1:
The patent replaces CPU-based software compression with dedicated hardware processing engines that perform video compression operations directly in hardware. This substitution of mechanical (software) processing with physical (hardware) processing significantly improves efficiency and reduces energy consumption.
Solution Approach 2:
The processing nodes are designed to autonomously perform compression operations on incoming video streams without requiring intensive CPU intervention. Each node self-manages its compression tasks, reducing the overall energy burden on the system's central processing units.
3Quantity of substance
If multiple camera sensors are deployed to capture the environment, then the system can gather comprehensive environmental data, but the data processing burden increases significantly
Solution Approach 1:
The patent assigns specific processing nodes to handle data from specific camera streams, dividing the overall processing burden into manageable segments. This allows the system to handle data from multiple cameras without proportionally increasing overall processing complexity.
Solution Approach 2:
The system merges the processing functions of multiple camera streams into a unified architecture where data from multiple cameras is consolidated and processed through coordinated nodes, reducing the cumulative complexity that would arise from completely independent processing paths.
Data Source
AI summary
A video system includes a first data gathering node configured to receive a plurality of image streams from a plurality of cameras, respectively. Each of the plurality of cameras captures an environment of an autonomous driving vehicle (ADV). The first data gathering node tags the plurality of image streams with metadata that identifies each of the plurality of image streams, and combines the plurality of image streams with the metadata to form a combined image stream. A second data gathering node is communicatively coupled to the first data gathering node and is to receive the combined image stream from the first data gathering node and output the combined image stream with a second combined image stream.