Distributed Image Cognition Processing for ADAS Object Tracking

Resolve Bottlenecks,
Find Innovative Solutions
Generate 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

VSEngineering 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

Engineering Contradiction:
Improveobject detection coverageVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

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.

Inventive Principle:
Principle #1Segmentation

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.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If multiple image sensors perform independent object detection and classification, then detection accuracy is improved, but computational complexity increases

Engineering Contradiction:
Improvedetection accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

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.

Inventive Principle:
Principle #5Merging (Combining)

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.

Inventive Principle:
Principle #26Copying

3Reliability

If image processors perform full object detection and classification, then processing completeness is improved, but processing overhead increases

Engineering Contradiction:
Improveprocessing completenessVSAvoidprocessing overhead
Core Design Contradiction:
ReliabilityVSLoss of energy

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.

Inventive Principle:
Principle #16Partial or excessive action

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.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3343505B1Distributed image cognition processing system
Publication Date: 2021.02.17 NXP USA INC
  • EP3343505B1 patent drawingFigure 1
  • EP3343505B1 patent drawingFigure 2
  • EP3343505B1 patent drawingFigure 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.