Adaptive Leaky Integrator for Periodic Light Artefact Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional leaky integrators fail to adequately suppress visible artefacts in images rendered on displays when capturing scenes with periodic lights, such as flashing indicators or overhead lighting, as they do not effectively dampen signals oscillating at frequencies around 1Hz.
Innovation Solution
A leaky integrator design that varies the dampening signal based on the presence, duty cycle, and frequency of periodic oscillations in the input signal, using a dampening signal calculated from upper and lower difference signals to selectively dampen periodic oscillations more than aperiodic ones, thereby reducing artefacts in final images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a conventional leaky integrator is used to filter signals, then sudden changes are avoided, but visible artefacts caused by periodic oscillations at 1Hz frequency are not adequately attenuated
Solution Approach 1:
The leaky integrator employs a dynamically adjustable dampening signal whose strength varies based on the detected characteristics of input signal oscillations. The dampening factor is modified in real-time according to the frequency and duty cycle of detected periodic oscillations, allowing the filter to adapt its behavior to different scene conditions rather than using a fixed dampening level.
Solution Approach 2:
The system changes the dampening parameter of the leaky integrator based on detected signal characteristics. By analyzing the frequency and duty cycle of oscillations in the input signal, the system adjusts the dampening factor to optimally suppress artefacts while preserving legitimate signal variations, thereby resolving the contradiction between smoothing and artefact reduction.
2Object-affected harmful factors
If the dampening signal is increased to suppress periodic oscillations, then visible artefacts are reduced, but legitimate signal changes may be overly dampened
Solution Approach 1:
The dampening signal is applied selectively based on the local characteristics of the input signal. By detecting specific patterns such as periodic oscillations at certain frequencies and duty cycles, the system applies stronger dampening only when artefacts are present, while maintaining normal signal processing when legitimate changes occur, thus preserving signal accuracy while reducing artefacts.
Solution Approach 2:
The system uses feedback from the input signal characteristics to dynamically adjust the dampening level. By continuously monitoring the frequency and duty cycle of oscillations, the system provides feedback to the leaky integrator to modulate the dampening factor, ensuring that dampening is applied appropriately only when needed to suppress artefacts without compromising legitimate signal information.
3Device complexity
If a fixed dampening factor is used in the leaky integrator, then the implementation is simple, but it cannot adapt to different frequencies and duty cycles of periodic oscillations
Solution Approach 1:
The leaky integrator transitions from a static design with fixed parameters to a dynamic design where the dampening factor is continuously adjusted based on detected signal characteristics. The system monitors the frequency and duty cycle of oscillations and dynamically modifies the dampening level accordingly, enabling adaptation to various artefact conditions while maintaining a relatively simple overall architecture.
Solution Approach 2:
The system changes the dampening parameter of the leaky integrator based on detected signal characteristics such as frequency and duty cycle of periodic oscillations. This parameter adaptation allows the integrator to effectively handle different types of artefacts without requiring a completely complex redesign, balancing simplicity with versatility.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to receiving an input signal (1002); generating (1004) an output signal by integrating a leaked signal over an integration time, wherein the leaked signal is obtained based on a dampening signal, a leak factor and the input signal; and providing (1012) the output signal.