Event Camera Optical Flow Tracking Without Image Conversion

Solution Overview

Problem

Existing object monitoring techniques using event cameras face inaccuracies and high computational costs due to the need to convert event camera data into images for processing.

Innovation Solution

Directly utilizing event camera data to determine temporally regularized optical flow velocities, allowing for accurate mapping of object movement without image conversion, using a computing device to process pixel events and apply a variational method to smooth optical flow velocities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If event camera data is converted into images for processing, then object monitoring can be performed using traditional computer vision techniques, but the measurement precision deteriorates and the computational complexity increases

Engineering Contradiction:

Improveobject monitoring capabilityVSAvoiddata accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The patent extracts and utilizes only the essential temporal and intensity change information from event camera data, bypassing the need for full image conversion. By focusing on optical flow computation directly from event streams, the method removes the unnecessary intermediate step of creating images, thereby preserving measurement precision while maintaining operational capability.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of converting event data to images and then analyzing movement (traditional approach), the patent inverts the process by directly computing optical flow from event camera data. This inversion eliminates the image conversion step that degrades precision and reduces computational complexity, while still enabling effective object monitoring.

Inventive Principle:

Principle #13The other way round (Inversion)

2Ease of operation

If event camera data is converted into images for processing, then traditional computer vision techniques can be applied, but the computational complexity increases

Engineering Contradiction:

Improveobject monitoring capabilityVSAvoidcomputational complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent extracts and utilizes only the essential temporal and intensity change information from event camera data, bypassing the need for full image conversion. By focusing on optical flow computation directly from event streams, the method removes the unnecessary intermediate step of creating images, thereby preserving measurement precision while maintaining operational capability.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of converting event data to images and then analyzing movement (traditional approach), the patent inverts the process by directly computing optical flow from event camera data. This inversion eliminates the image conversion step that degrades precision and reduces computational complexity, while still enabling effective object monitoring.

Inventive Principle:

Principle #13The other way round (Inversion)

3Measurement precision

If temporally regularized optical flow velocities are determined directly from event camera data, then measurement precision improves, but the algorithmic complexity increases

Engineering Contradiction:

Improveoptical flow accuracyVSAvoidprocessing algorithm complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies temporal regularization to the optical flow velocity computation, which introduces a smoothing parameter that balances accuracy with computational feasibility. By regularizing the optimization problem, the method achieves higher measurement precision while controlling algorithmic complexity through parameter-based constraints rather than more complex algorithms.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent computes optical flow continuously over time using temporal regularization, which leverages information from multiple time steps to improve accuracy. This continuous approach maintains precision by incorporating temporal coherence while managing complexity through efficient recursive computation that builds on previous frames rather than reprocessing all data.

Inventive Principle:

Principle #20Continuity of useful action

Data Source

PatentUS20240177484A1Object monitoring using event camera data

Publication Date: 2024.05.30 ROSEMOUNT AEROSPACE INC

AI summary

A method includes obtaining data from an event camera for each of a plurality of time instances. The data includes events corresponding to changes detected by a corresponding plurality of pixels of the event camera at each time instance. Temporally regularized optical flow velocities are determined at each time instance. Each of the pixels has a respective one of the optical flow velocities at each time instance. An optical flow of a feature of an object in a field of view of the event camera is determined based on a predetermined relationship between the temporally regularized optical flow velocities at a selected time instance and the temporally regularized optical flow velocities at a subsequent time instance. The feature of the object corresponds to one of the plurality of events at the selected time instance and one of the plurality of events at the subsequent time instance.