Event Sensor Motion Vector Calculation with Overlapping Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion vector calculation methods in event-based vision sensors are limited by a fixed mapping time, which affects the precision of motion vector calculation and is not adaptable to varying motion vector magnitudes or subject shapes, leading to suboptimal calculation periods.
Innovation Solution
A motion vector calculation device that overlaps the mapping of event data across multiple frames, allowing for partial overlap of first time periods to generate frames, thereby improving the calculation period and accuracy of motion vectors based on a frame group with varying start times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed mapping time is used for generating frames, then the device complexity is reduced, but the measurement precision of motion vectors deteriorates
Solution Approach 1:
The patent applies dynamics by making the mapping time variable rather than fixed. The control unit dynamically adjusts the mapping time for each frame based on the magnitude of motion vectors detected in previous frames. When motion vector magnitude is large, the mapping time is extended to maintain precision; when motion vector magnitude is small, the mapping time is reduced to improve processing efficiency. This dynamic adaptation resolves the contradiction between device complexity and measurement precision.
2Measurement precision
If the mapping time is extended to improve motion vector precision, then the measurement precision is improved, but the productivity deteriorates
Solution Approach 1:
The system dynamically adjusts mapping time based on motion vector magnitude. For small motion vectors, a shorter mapping time is used to maintain high processing speed. For large motion vectors, the mapping time is extended to ensure accurate measurement. This dynamic adjustment resolves the contradiction between precision and productivity by adapting the mapping time to the actual motion characteristics of the scene.
Solution Approach 2:
The patent changes the parameter of mapping time based on detected motion vector magnitudes. By monitoring the magnitude of motion vectors and adjusting the mapping time parameter accordingly, the system optimizes both precision and processing speed. This parameter adaptation allows the system to maintain high productivity for static or slowly moving scenes while achieving high precision for fast-moving scenes.
3Productivity
If the mapping time is reduced to improve processing speed, then the productivity is improved, but the measurement precision deteriorates
Solution Approach 1:
The system employs dynamic mapping time adjustment where the mapping time is reduced only when motion vector magnitudes are small, maintaining high processing speed. When large motion vectors are detected, the mapping time is automatically extended to preserve measurement precision. This dynamic behavior resolves the contradiction by adapting the mapping time to actual scene requirements rather than using a fixed value.
Data Source
AI summary
An optical device acquires event data based on an output of an event sensor detecting a change in luminance of a subject image and maps the event data acquired in a mapping time to generate a frame. The optical device performs control such that the mapping on the event data is overlapped partially in a plurality of the frames and calculates a motion vector based on the plurality of frames in which there is a difference of the mapping time at a start time of the mapping.