Phase-Detection Auto Focus Using Sub-Region Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Phase detection auto focus methods in existing technologies are limited by the need for shading pixels, which are affected by noise and light intensity, leading to inaccurate phase values, prolonged focus time, and reduced precision.
Innovation Solution
The method divides a focus region into sub-regions, calculates phase differences and confidence levels, performs dynamic matching in weight tables to obtain weight values, and performs weighted summation to improve focus precision and reduce focus time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If phase detection auto focus uses shading pixels to determine focus offset, then focusing speed is improved, but focus precision deteriorates due to noise and light intensity affecting phase values
Solution Approach 1:
The focus region is divided into multiple sub-regions, and each sub-region is independently processed to calculate phase difference and confidence level. This segmentation allows selective weighting of reliable sub-region data, improving overall focus precision while maintaining speed through parallel processing of divided regions.
Solution Approach 2:
Different sub-regions are assigned different weight values based on their local confidence levels and depth of field characteristics. High-confidence sub-regions with suitable depth of field contribute more to the final focus determination, while low-confidence or unsuitable sub-regions are downweighted or excluded, thereby improving measurement precision without sacrificing processing speed.
2Speed
If simple blocking device is added in imaging optical path, then rapid phase detection auto focus is achieved, but device complexity increases
Solution Approach 1:
The blocking device is designed to serve multiple functions: it creates the necessary phase shift for focus detection while also enabling the system to differentiate between suitable and unsuitable sub-regions based on depth of field. This multi-functionality reduces the need for additional separate components, thereby limiting the increase in device complexity.
Solution Approach 2:
The system uses the phase detection data from the blocking device to automatically identify and weight suitable sub-regions without requiring external intervention or additional hardware. The algorithm self-adjusts weights based on confidence levels and depth of field analysis, reducing the need for complex external control mechanisms.
3Ease of manufacture
If all sub-regions are processed equally, then processing is simplified, but focus precision deteriorates due to inclusion of low-confidence data
Solution Approach 1:
The weight values for different sub-regions are dynamically adjusted based on their confidence levels and depth of field suitability. Instead of static equal weighting, the system adaptively assigns higher weights to high-confidence sub-regions and lower or zero weights to low-confidence ones, improving measurement precision while maintaining reasonable processing complexity through algorithmic adaptation.
Solution Approach 2:
The system changes the parameter of weight values from a fixed equal distribution to a variable distribution based on confidence levels and depth of field characteristics. This parameter transformation allows the system to emphasize reliable data and suppress unreliable data, thereby improving focus precision without requiring fundamentally more complex processing architecture.
Data Source
AI summary
A phase-detection auto focus method includes: dividing a focus region into a plurality of sub-regions, and respectively calculating a phase difference and a confidence level of each sub-region; performing dynamic matching in a corresponding weight table according to the phase difference and the confidence level of each sub-region, and performing a table lookup to obtain a corresponding weight value; calculating a phase difference and a confidence level of the focus region in conjunction with the weight value; and performing auto focusing according to the phase difference and the confidence level of the focus region.