ADAS Sensor Data Decompression for Lossy Compression Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ADAS systems face challenges in reducing the data quantity of sensor data for transmission and storage, with limited compression rates using lossless algorithms, leading to high memory and bandwidth requirements.
Innovation Solution
Implementing lossy compression algorithms and architectures for ADAS sensor data processing units and systems, which include an input interface, decompression module, processing unit, and output unit, to achieve higher compression rates, and integrating compression and decompression modules within ADAS sensors or as separate units for efficient data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If lossless compression algorithms are used to reduce data quantity, then data integrity is maintained, but compression rates are limited
Solution Approach 1:
The patent changes the compression parameter from lossless to lossy compression, accepting controlled information loss in exchange for significantly higher compression rates. This is achieved through configurable compression strengths that allow balancing between data quality and compression efficiency.
Solution Approach 2:
The system discards redundant or less important sensor data information during compression, and recovers essential information through selective decompression and processing. The decompression module reconstructs data with acceptable quality for ADAS processing while achieving high compression ratios.
2Productivity
If lossy compression is used to achieve higher compression rates, then data quantity is reduced, but data quality deteriorates
Solution Approach 1:
The system applies different quality levels to different parts of the sensor data based on their importance for ADAS functions. Critical regions receive higher quality preservation while less critical areas use higher compression, achieving overall data quality maintenance with improved compression rates.
Solution Approach 2:
The system applies partial decompression or selective reconstruction of compressed data, processing only the most essential information at full quality while accepting reduced quality for less critical data, thereby balancing compression rate and data quality.
3Measurement precision
If high-resolution sensor data is captured to improve detection range and object resolution, then detection capability is enhanced, but data quantity increases
Solution Approach 1:
The system extracts and processes only the most relevant information from high-resolution sensor data, such as detecting objects, lane markings, and traffic signs, while compressing and storing less critical data at lower resolutions, thereby maintaining detection capability with reduced data quantity.
Solution Approach 2:
The sensor data is segmented into different regions of interest based on their importance for ADAS functions. Critical regions are processed at full resolution while other regions are compressed more aggressively, enabling selective preservation of measurement precision where needed.
4Reliability
If algorithms are validated on lossless data to ensure accuracy, then algorithm reliability is maintained, but validation complexity increases for lossy data
Solution Approach 1:
The system performs preliminary validation of algorithms on lossless reference data to establish baseline performance, then applies the same algorithms to lossy-compressed data with adjusted expectations. This preliminary validation approach simplifies the overall validation process by creating a reference framework.
Solution Approach 2:
The system creates reference copies of ground truth data and compares algorithm outputs against these references both for lossless and lossy-compressed data. This copying and comparison approach provides a systematic method for validating algorithm reliability across different compression levels.
Data Source
AI summary
Example embodiments relate to an ADAS sensor data processing unit, to an ADAS sensor system and to an ADAS sensor data evaluation method for use in driver assistance systems or systems for the automated driving of a vehicle. The ADAS sensor data processing unit includes an input interface, a decompression module, a processing unit and an output unit. The input interface is designed to receive data of an ADAS sensor that have been subjected to lossy compression by a compression module. The decompression module is designed to decompress the compressed data of the ADAS sensor. The processing unit is designed to process the decompressed data (IdSD) of the ADAS sensor, information relevant to an ADAS/AD function being ascertained from the decompressed sensor data. The output unit is designed to output the ascertained information relevant to the ADAS function.