Event-to-Frame Interface for Dynamic Vision Sensor Compatibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in interfacing event-based processing systems with frame-based processing systems, as the output from event-based object detection systems, such as those using dynamic vision sensors, is not compatible with the input requirements of frame-based processing systems, leading to inefficiencies in processing and classification of detected objects.
Innovation Solution
A method and apparatus that convert multiple events from an event sensor into a frame, which can be input into a frame-based processing system, allowing for real-time processing and classification, with optional filtering to manage processing load and noise reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If event-based processing system output is directly used, then real-time detection capability is maintained, but compatibility with frame-based processing systems deteriorates
Solution Approach 1:
The patent introduces an interface module as an intermediary between the event-based processing system and frame-based processing system. This interface converts asynchronous event streams into synchronous frame-based data structures, enabling compatibility while preserving the real-time advantages of event-based detection. The intermediary translates data formats and timing mechanisms without losing the speed benefits.
2Adaptability or versatility
If all events are converted to frames, then compatibility with frame-based systems is improved, but processing load increases
Solution Approach 1:
The patent extracts and processes only the most relevant events for conversion to frame-based format, rather than converting all events. By filtering and selecting significant events based on detection criteria, the system maintains compatibility with frame-based systems while reducing the processing load to manageable levels.
3Productivity
If event filtering is applied, then processing efficiency is improved, but noise detection capability may deteriorate
Solution Approach 1:
The patent applies different filtering criteria to different regions or types of events. By customizing filtering thresholds and parameters based on local characteristics of event streams, the system maintains high processing efficiency for obvious signals while preserving the ability to detect subtle noise events that require lower thresholds or different filtering approaches.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method of interfacing an event based processing system with a frame based processing system is presented. The method includes converting multiple events into a frame. The events may be generated from an event sensor. The method also includes inputting the frame into the frame based processing system.