Event Camera Asynchronous Sensor for Power-Efficient Image Capture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional cameras require high power and memory due to synchronous data processing of light-ray parameters, and generate unnecessary data for static pixels, leading to inefficient image capture.
Innovation Solution
An event camera with an event sensor that operates in asynchronous mode, outputting data only when intensity changes exceed a threshold, allowing for faster data output and reduced power consumption by generating images based on relative intensity changes rather than static frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional cameras integrate light-ray parameters over a specified period of time in synchronous mode, then complete image frames can be captured, but power consumption and memory requirements increase heavily
Solution Approach 1:
The event camera operates by periodically evaluating intensity changes at each pixel against a threshold, triggering asynchronous events only when changes occur. This periodic evaluation replaces continuous synchronous integration, reducing power consumption while maintaining image capture capability through event-based representation
Solution Approach 2:
The system extracts only the necessary information (intensity changes exceeding threshold) from the full image data. By outputting only changed pixels as events rather than complete frames, the system reduces data volume and power consumption while preserving essential visual information for dynamic scenes
2Loss of information
If traditional cameras read out data for every photodiode in synchronous manner, then complete image data is obtained, but data throughput is limited and unnecessary data is generated for static pixels
Solution Approach 1:
The system extracts only changed pixel information from the sensor array, outputting events solely for pixels where intensity changes exceed the threshold. This eliminates data transmission for static pixels while preserving complete information about dynamic regions, significantly increasing effective data throughput
Solution Approach 2:
The camera transitions from static synchronous frame capture to dynamic asynchronous event streaming. The output rate adapts to scene activity, with data throughput automatically increasing for dynamic scenes and decreasing for static scenes, optimizing bandwidth utilization while maintaining information completeness
3Duration of action of stationary object
If traditional cameras capture full image frames at fixed intervals, then temporal coverage is ensured, but memory requirements increase for storing and processing large volume of data
Solution Approach 1:
The system extracts and stores only event data representing intensity changes over time, rather than storing complete frames. This reduces memory requirements from storing full-resolution images to storing sparse event streams, while maintaining temporal coverage through continuous event recording
Solution Approach 2:
The system discards redundant information from static pixels that provide no useful temporal data. By recording only changes, the system recovers efficient storage while preserving all dynamically relevant information, allowing long-term temporal coverage with minimal memory usage
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 event camera achieves efficient image generation with lower power and memory usage, improving data throughput and reducing unnecessary data generation for static pixels, enabling faster and more efficient image capture.
Implementation Method 1
The event sensor includes a plurality of photodiodes and is configured to operate in one or modes including an event mode. An event sensor operating in the event mode asynchronously outputs data values corresponding to relative intensity changes within the local area.
Data Source
AI summary
An imaging device operates as an event camera. The device includes an event sensor and a controller. The sensor comprises a plurality of photodiodes that asynchronously output data values corresponding to relative intensity changes within a local area. The controller populates an event matrix based in part on data values asynchronously received from the sensor and positions of photodiodes associated with the received data values over a first time period. The controller populates a change matrix based in part on a threshold intensity value and the photodiodes associated with the received data values over the first time period, and generates an image for the first time period using the event matrix and the change matrix.