Imaging Lens Autofocus Accuracy via Threshold Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autofocus methods struggle to accurately detect the in-focus position of an imaging lens, especially when the subject is extremely out of focus, leading to inaccurate focusing even when using skin-tone areas or contrast information.
Innovation Solution
An apparatus and method involving a solid-state electronic image sensing device, a focusing data output device, a target-image detecting device, and an in-focus position deciding device, which moves the imaging lens in increments along the optic axis, detects targets at positions with sufficient focusing data, and decides the in-focus position based on detected data, using prescribed distances and threshold values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If skin-tone area detection is used as a focusing area, then focusing can be achieved, but accurate detection becomes difficult when the subject is extremely out of focus
Solution Approach 1:
The patent performs preliminary focus adjustment by moving the lens to a position where focus evaluation data meets a threshold before attempting target detection. This preliminary action ensures the subject is sufficiently in focus, making subsequent target detection (skin-tone or contrast-based) accurate and reliable.
Solution Approach 2:
The patent dynamically adjusts the lens position based on focus evaluation data, moving the lens incrementally until the focus metric exceeds a threshold. This dynamic adjustment ensures optimal focus conditions are achieved before target detection begins, adapting to different subject distances and focus conditions.
2Reliability
If contrast information alone is used for target detection, then focusing can be attempted, but accurate target detection fails when the subject is extremely out of focus
Solution Approach 1:
The patent performs preliminary focus adjustment by moving the lens to a position where focus evaluation data meets a threshold before attempting target detection. This preliminary action ensures the subject is sufficiently in focus, making subsequent target detection (skin-tone or contrast-based) accurate and reliable.
Solution Approach 2:
The patent dynamically adjusts the lens position based on focus evaluation data, moving the lens incrementally until the focus metric exceeds a threshold. This dynamic adjustment ensures optimal focus conditions are achieved before target detection begins, adapting to different subject distances and focus conditions.
3Speed
If the lens is moved in large increments, then focusing speed is improved, but focusing precision deteriorates
Solution Approach 1:
The patent segments the focusing process into two phases: a coarse search phase using larger increments to quickly locate the focus range, and a fine adjustment phase using smaller increments to precisely determine the in-focus position. This segmentation allows the system to achieve both speed and precision.
Solution Approach 2:
The patent dynamically adjusts the lens position based on focus evaluation data, moving the lens incrementally until the focus metric exceeds a threshold. This dynamic adjustment ensures optimal focus conditions are achieved before target detection begins, adapting to different subject distances and focus conditions.
Data Source
AI summary
A subject is imaged while an imaging lens is moved along the optic axis in incremental distances, and a graph of focus evaluations representing degree of focus of the subject image is obtained. Processing for detecting a face is executed at positions of the imaging lens where the focus evaluation value is equal to or greater than a threshold value. If a face is detected from the image of the subject obtained at positions of the imaging lens, then this range of positions is set as a focusing range and the subject is imaged while the imaging lens is moved within this focusing range. A graph of face-area focus evaluations representing degree of focus of the face detected from among the obtained images of the subject is obtained and the position at which this graph is maximum is decided upon as the in-focus position of the imaging lens.