Unbiased Focus Metric Calculation Using Image Shifting and Temporal Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automatic focus control algorithms in cameras are negatively biased by noise and illumination variations, leading to suboptimal focus positions, especially under non-ideal conditions such as fluctuating illumination and high noise levels.
Innovation Solution
A method and apparatus for calculating an unbiased focus metric by processing a focusing image and its shifted version, which is insensitive to noise and illumination changes, using gradient multiplication and normalization to determine the proximity to optimal focus without requiring de-noising steps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional FOM calculations are used with edge detection algorithms, then focus position can be determined under ideal conditions, but the algorithm fails under non-ideal conditions with noise and illumination variations
Solution Approach 1:
The patent converts the harmful effects of noise and illumination variations into a beneficial filtering mechanism. By using a temporal low-pass filter on the FOM values, the algorithm allows rapid response to genuine focus changes while automatically suppressing spurious variations caused by noise and lighting fluctuations. This transforms the previously harmful high-sensitivity characteristic into a beneficial feature for distinguishing real focus events from noise.
Solution Approach 2:
The patent introduces a new parameter - the temporal low-pass filter with adjustable time constant - to modify the FOM calculation behavior. By changing the filtering parameter, the system can adapt to different noise levels and illumination conditions, making the focus algorithm reliable across varying environmental conditions while maintaining measurement precision.
2Productivity
If focusing images are captured over a short period to minimize focusing time, then productivity is improved, but noise level increases leading to biased FOM calculations
Solution Approach 1:
The patent converts the harmful effect of high noise levels (resulting from short capture periods) into a beneficial feature. The temporal low-pass filter exploits the fact that noise varies rapidly while genuine focus changes are more sustained. By filtering at an appropriate time constant, the system achieves accurate FOM calculations even with noisy, rapidly-captured images, thus maintaining both productivity and measurement precision.
3Ease of operation
If the focus control algorithm responds to FOM changes, then focus adjustment is made, but erroneous corrections occur when FOM changes are caused by illumination variations rather than focus errors
Solution Approach 1:
The patent converts the harmful effect of illumination-induced FOM variations into a beneficial filtering opportunity. By applying the temporal low-pass filter, the system distinguishes between rapid FOM changes caused by illumination fluctuations (filtered out) and sustained changes indicating genuine focus errors (passed through). This enables accurate automatic focus control without erroneous corrections.
Solution Approach 2:
The temporal low-pass filter acts as an intermediary between the raw FOM calculation and the focus control decision. This intermediary component processes the FOM signal to remove spurious variations from illumination changes while preserving genuine focus error signals, thereby improving focus control accuracy without compromising ease of operation.
Data Source
AI summary
A method comprises receiving a focusing image and shifting the focusing image to obtain a shifted focusing image. In the method, a focus metric of the focusing image is calculated from the focusing image and the shifted focusing image, where the focus metric is configured for use as a factor in making an automatic focus adjustment determination. An automatic focusing apparatus includes a controller configured to implement the method.