Event-Driven Vision and IMU Coordinate Conversion for Low-Latency User Display
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods have not sufficiently explored the combination of event-driven type vision sensors with other devices to generate precise user condition displays while minimizing latency.
Innovation Solution
A system utilizing a first image sensor for synchronous pixel scanning, an event-driven type vision sensor for asynchronous light intensity detection, an inertial sensor for attitude information, and computation processing devices for coordinate conversion and image generation, to create a display image reflecting user conditions with precision and reduced latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If event-driven type vision sensors are used to detect changes in light intensity asynchronously, then power consumption is reduced and processing speed is improved, but coordinate conversion accuracy and system integration complexity increase
Solution Approach 1:
The system divides the vision sensor into multiple pixel units, each independently detecting light intensity changes and generating event signals. This segmentation allows asynchronous processing while maintaining coordinate accuracy through per-pixel event detection and subsequent spatial coordination.
Solution Approach 2:
The patent introduces an intermediary coordinate conversion process that mediates between the asynchronous event-driven sensor data and the synchronous display system. This intermediary layer handles the complexity of coordinate transformation and synchronization, isolating the event-driven processing from the display pipeline.
2Measurement precision
If coordinate conversion is performed on event-driven image signals, then display precision is improved, but processing time and latency increase
Solution Approach 1:
The system performs preliminary coordinate conversion calculations based on pre-acquired attitude information from inertial sensors. By preparing conversion parameters in advance and updating them continuously, the actual image processing can proceed faster with reduced latency while maintaining precision.
Solution Approach 2:
The patent implements feedback mechanisms where attitude information from inertial sensors is continuously updated and fed back into the coordinate conversion process. This feedback loop ensures that coordinate transformations remain accurate despite changes in sensor orientation, reducing the need for reprocessing.
3Measurement precision
If multiple image sensors are used for comprehensive user detection, then detection accuracy is improved, but communication load and data processing requirements increase
Solution Approach 1:
The system extracts only the essential event signals from the image sensors - specifically, changes in light intensity at individual pixels - rather than processing complete image frames. This extraction approach maintains detection accuracy for user position and movement while significantly reducing the quantity of data that needs to be processed and transmitted.
Solution Approach 2:
The patent applies partial action by processing only the necessary portions of image data - specifically, pixel-level event changes rather than entire images. This selective processing maintains sufficient detection accuracy while reducing overall data volume and communication load.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system generates a display image that accurately reflects user conditions by calculating and converting coordinate information, suppressing latency and improving installation flexibility while reducing communication load and privacy concerns.
Implementation Method 1
an event-driven type vision sensor that, upon detecting a change in an intensity of incident light on each of the pixels, generates a second image signal asynchronously
Data Source
AI summary
A system includes a first image sensor that generates a first image signal by synchronously scanning all pixels at a prescribed timing, a second image sensor including an event-driven type vision sensor that, upon detecting a change in an intensity of incident light on each pixel, generates a second image signal asynchronously, an inertial sensor that acquires attitude information on the first image sensor and the second image sensor, a first computation processing device that recognizes a user on the basis of at least the second image signal and calculates coordinate information regarding the user on the basis of at least the second image signal, a second computation processing device that performs coordinate conversion on the coordinate information on the basis of the attitude information, and an image generation device that generates a display image which indicates a condition of the user, on the basis of the converted coordinate information.