Autofocus Area Sizing for Subject Attribute Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging technologies face challenges in achieving accurate focusing, particularly due to the inclusion of background images in autofocus target regions and erroneous subject tracking, which affect the precision of focusing and tracking operations.
Innovation Solution
An imaging method and apparatus that utilize a machine-learned model to detect a subject range, determine its attribute, and adjust the size of a second range for acquiring distance information based on the attribute, allowing for improved focusing accuracy by selectively varying the size of the AF area and correcting it according to the subject's state and reliability of determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the AF target region is expanded to include more background area for distance information acquisition, then the distance measurement capability is improved, but the focusing accuracy deteriorates due to inclusion of non-subject regions
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different regions within the detection range. The AF area is selectively adjusted based on subject attributes - for moving objects like vehicles and animals, the AF area is expanded beyond the subject boundary to capture background distance information, while for static objects like buildings, the AF area is restricted to the subject boundary. This regional differentiation resolves the contradiction by applying different quality standards to different parts of the imaging scene.
Solution Approach 2:
The patent implements dynamics by making the AF area size adaptive rather than fixed. The system dynamically adjusts the AF area based on real-time subject detection results and attribute classification. When a moving object is detected, the AF area automatically expands to include background regions for distance information; when a static object is detected, the AF area contracts to maintain focusing precision. This dynamic adjustment mechanism allows the system to optimize both distance measurement and focusing accuracy according to the specific imaging scenario.
2Manufacturing precision
If the AF area is restricted to within the subject range to improve focusing precision, then the focusing accuracy is improved, but the distance information acquisition capability deteriorates
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different regions within the detection range. The AF area is selectively adjusted based on subject attributes - for moving objects like vehicles and animals, the AF area is expanded beyond the subject boundary to capture background distance information, while for static objects like buildings, the AF area is restricted to the subject boundary. This regional differentiation resolves the contradiction by applying different quality standards to different parts of the imaging scene.
Solution Approach 2:
The patent implements dynamics by making the AF area size adaptive rather than fixed. The system dynamically adjusts the AF area based on real-time subject detection results and attribute classification. When a moving object is detected, the AF area automatically expands to include background regions for distance information; when a static object is detected, the AF area contracts to maintain focusing precision. This dynamic adjustment mechanism allows the system to optimize both distance measurement and focusing accuracy according to the specific imaging scenario.
3Reliability
If the subject detection range is expanded to capture more context, then the subject identification capability is improved, but the processing complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the complex subject detection and classification task into distinct functional modules. The detection unit identifies the subject range, the attribute determination unit classifies the subject type (moving vs. static), and the control unit adjusts the AF area accordingly. This modular segmentation reduces processing complexity by breaking down the overall task into manageable stages, each handling a specific aspect of subject analysis.
Solution Approach 2:
The patent extracts only the essential attribute information needed for AF area control - specifically, whether the subject is a moving object or a static object. Rather than performing comprehensive subject analysis, the system extracts and utilizes only the critical movement status attribute. This extraction approach simplifies processing by focusing on the minimal necessary information for achieving the desired focusing behavior.
Data Source
AI summary
There is provided an imaging method including: an imaging step of generating image data by an imaging element; a detection step of detecting a first range including a subject that is a focusing target from the image data; a determination step of determining an attribute of the subject; and a decision step of deciding whether a size of a second range for acquiring distance information of the subject is set to be within the first range or to exceed the first range, based on the attribute.