Hybrid Auto Focus Tracking Method for Video Stability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Phase detection image sensors face challenges in maintaining stable and reliable video focus tracking in complex environments with varying lighting and textures.
Innovation Solution
A hybrid auto focus tracking method that determines a first lens focusing position based on phase difference and a second lens focusing position based on zoom ratio of feature points within a region of interest, then fuses these positions to achieve a final lens focusing position for improved focus tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If phase detection image sensors are used for auto focus, then focusing speed and accuracy are improved, but stability and reliability in complex environments deteriorate
Solution Approach 1:
The patent combines phase detection autofocus with feature point-based autofocus into a hybrid system. The processor determines a first lens focusing position based on phase difference and a second lens focusing position based on zoom ratio of feature points, then fuses these positions to obtain a final lens focusing position. This merging of two different focusing methodologies allows the system to maintain fast focusing while improving reliability in complex environments by cross-validating results from multiple approaches.
2Reliability
If phase detection image sensors work in complex environments, then focusing capability is maintained, but stability of video focus tracking deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where the processor continuously monitors the zoom ratio of feature points within a region of interest and adjusts the lens focusing position accordingly. By comparing the current feature point positions with previous frames and calculating the zoom ratio, the system can detect when the subject is moving or changing scale and dynamically adjust focus to maintain stable tracking throughout the video sequence.
Solution Approach 2:
The hybrid approach merges phase detection results with feature point tracking results. The processor fuses the first lens focusing position (from phase detection) and the second lens focusing position (from feature point zoom ratio) to generate a final focusing position. This combination allows the system to leverage the speed advantage of phase detection while using feature point analysis to ensure stability in complex environments with varying lighting and textures.
Data Source
AI summary
A hybrid auto focus tracking method and a hybrid auto focus tracking device are provided. The hybrid auto focus tracking method comprises: determining a first lens focusing position based on a phase difference of a phase detection image output by a phase detection image sensor; determining a second lens focusing position based on a zoom ratio of feature points within a region of interest in a scene image output by the phase detection image sensor; determining a final lens focusing position by fusing the first lens focusing position and the second lens focusing position; and performing focus tracking of the phase detection image sensor based on the final lens focusing position.