Subject Region Selection via Motion Vector Prioritization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image capture apparatuses struggle to accurately determine the main subject in group motion scenes, such as sports, where multiple subjects move in the same direction, often capturing a range different from the user's intended view.
Innovation Solution
An image processing apparatus that detects subject regions, obtains motion vectors for each region, and determines the main subject based on a prioritized direction of movement set in advance, allowing for appropriate subject selection and control of the image capture unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the subject region located at the beginning of the direction of movement is selected as the main subject region, then the main subject region can be determined in group motion scenes, but the captured range may not match the user's intended view in sports where multiple subjects move in the same direction
Solution Approach 1:
The patent applies dynamics by making the main subject determination criteria adaptable to different shooting scenes. The determination unit dynamically selects determination criteria based on the detected scene type (e.g., sports, street photography, concerts). For sports scenes, it uses criteria that consider motion vectors and directional movement patterns, while for other scenes it may use different criteria such as subject size or position. This dynamic adaptation resolves the contradiction by allowing accurate main subject determination in each specific scene context rather than using a single fixed criterion.
Solution Approach 2:
The patent changes parameters based on shooting scene type. The determination unit modifies the determination criteria parameters according to the detected scene. In sports scenes, it prioritizes motion vector analysis and directional movement patterns, while in other scenes it may prioritize different parameters such as subject position or size. This parameter change approach enables the system to accurately determine main subjects across diverse shooting scenarios, resolving the contradiction between precision in specific scenes and versatility across all scenes.
2Device complexity
If a fixed criterion is used to determine the main subject region, then the determination process is simple, but it cannot accurately identify the main subject in various types of shooting scenes
Solution Approach 1:
The system dynamically adjusts the determination criteria based on the detected shooting scene type. The determination unit switches between different determination methods depending on whether the scene is sports, street photography, concerts, or other types. This dynamic approach maintains reasonable complexity by only selecting from pre-defined criteria sets, while significantly improving accuracy for each specific scene type compared to a single fixed criterion.
Solution Approach 2:
The patent creates multiple determination criteria sets (copies of determination logic) tailored to different shooting scenes. Each scene type has its own optimized determination criteria stored and selected as needed. This copying approach allows the system to maintain simple, scene-specific determination processes while achieving high accuracy across diverse shooting scenarios, resolving the contradiction between complexity and precision.
Data Source
AI summary
Disclosed is an image processing apparatus that detects one or more subject regions from an image. The image processing apparatus obtains a motion vector for each of the one or more subject regions and determines, as a main subject region, a subject region, among the one or more subject regions, for which a direction indicated by the motion vector is a prioritized direction of movement determined in advance.