Dynamic Diffuser Camera for Motion Blur Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cameras face challenges in reducing motion blur, especially under low light conditions, as they require long exposure times to capture sufficient light, leading to blurred images of moving objects and limited dynamic range, and existing methods for motion blur reduction are either inaccurate or impractical for real-time surveillance and snapshot photography.
Innovation Solution
The introduction of a dynamic diffuser in the camera's light path, which modulates its diffusing properties during exposure to counteract motion blur by using an inverse point spread function (IPSF) to deconvolute the image, allowing for motion-invariant imaging without the need for precise motion estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If long exposure times are used to capture sufficient light under low light conditions, then signal-to-noise ratio is improved, but motion blur increases
Solution Approach 1:
The patent applies a static diffuser that creates dynamic blur patterns during exposure. The diffuser remains fixed in position but its light-scattering effect creates time-varying blur kernels as objects move through the scene, enabling motion blur reduction while maintaining long exposure times for adequate signal-to-noise ratio.
Solution Approach 2:
The patent changes the optical parameters by introducing a diffuser with specific scattering properties. This transforms the point spread function into a motion-invariant blur kernel, allowing the system to capture sufficient light over long exposure times while the deconvolution process recovers sharp images by reversing the controlled blur effect.
2Manufacturing precision
If conventional motion blur reduction methods are used, then image sharpness is improved, but system complexity or practicality deteriorates
Solution Approach 1:
The patent uses a simple, inexpensive static diffuser element that can be easily integrated into the optical path. This single-component solution replaces complex active systems such as moving lenses, variable aperture mechanisms, or computational motion estimation algorithms, significantly reducing system complexity while achieving motion blur reduction.
3Measurement precision
If aperture is increased to improve depth sensitivity in depth-from-defocus, then depth estimation precision is improved, but motion blur increases
Solution Approach 1:
The patent segments the blur formation process into controlled diffuser scattering and object motion components. By using a static diffuser with known scattering properties, the system creates predictable blur kernels that can be separated from motion effects through deconvolution, enabling both depth estimation and motion blur reduction to proceed simultaneously without mutual interference.
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
This approach enables the capture of sharp images with high signal-to-noise ratio even under challenging conditions, such as low light and varying object velocities, without the need for motion estimation, effectively reducing motion blur and increasing the depth of field.
Implementation Method 1
a diffuser (4) is present in the light path between the lens (2) and the image sensor (3)... the more the diffusing properties of the diffuser (4) are modulated during the image integration
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and camera wherein the camera comprises in the light path a diffuser (4). The system or camera comprises a means (6) to modulate the diffusing properties of the diffuser (4) on an image projected by the lens on the sensor during exposure of the image. To the captured blurred image (10) an inverse point spread function is applied to deconvolute (24) the blurred image to a sharper image. Motion invariant image can so be achieved.