Distributed Image Cognition Processing for ADAS Object Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced driver assistance systems (ADAS) in vehicles face computational intensity due to duplicative object detection and classification by multiple image sensors, which can lead to delayed object detection and compromised driver safety.
Innovation Solution
A distributed image cognition processing system where image cognition processors share object metadata, reducing duplicative processing by having one processor track and classify objects and sending metadata to adjacent processors, allowing them to focus on object tracking rather than detection and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple image sensors perform independent object detection and classification, then object detection coverage is improved, but processing time increases and detection speed decreases
Solution Approach 1:
The patent segments the object processing workflow into distinct functional modules: object detection, object classification, and object tracking. Different image sensors can perform detection independently while sharing classification and tracking results through the segmented architecture, reducing redundant processing time while maintaining comprehensive detection coverage.
Solution Approach 2:
The system performs preliminary object detection by multiple sensors simultaneously, then shares detection results before proceeding to classification and tracking. This preliminary action allows parallel processing of detection while subsequent steps benefit from shared information, reducing overall processing time.
2Measurement precision
If multiple image sensors perform independent object detection and classification, then detection accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent merges the classification and tracking functions across multiple image sensors, so that once an object is detected by any sensor, its classification and tracking are performed centrally or shared among sensors. This combining approach maintains high detection accuracy through multiple sensors while significantly reducing the computational complexity of repeated classification operations.
Solution Approach 2:
The system creates copies of detection results and shares them across the sensor network. Instead of each sensor independently performing full detection, classification, and tracking, detection results are copied and distributed, allowing sensors to focus on tracking known objects rather than redundant detection and classification.
3Reliability
If image processors perform full object detection and classification, then processing completeness is improved, but processing overhead increases
Solution Approach 1:
The patent applies partial action by having image sensors perform only object detection initially, while object classification and tracking are performed selectively based on shared results. This partial execution of the full processing pipeline maintains processing completeness for critical functions while reducing overall processing overhead by avoiding redundant operations.
Solution Approach 2:
The system implements self-service through automated result sharing and coordination between sensors. Once one sensor detects and classifies an object, the results are automatically shared with other sensors, which then use this information for tracking without needing to perform independent classification, reducing processing overhead while maintaining completeness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Embodiments of an image cognition processing system are provided, including a method that includes detecting an object in a first field of view of a first image sensor, wherein the first image sensor is coupled to a first image cognition processor; generating tracking metadata for the object, wherein the generating is performed by the first image cognition processor, and the tracking metadata describes movement of the object; determining that the object is moving toward a second field of view of a second image sensor, wherein the second image sensor is located adjacent to the first image sensor; and providing the tracking metadata for the object to a second image cognition processor coupled to the second image sensor.