Image Processing Subject Region Extension for Detection Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques struggle to accurately select a desired subject for image capture when non-detection parts are present among multiple detected parts, leading to potential misselection.
Innovation Solution
An image processing device that detects multiple parts related to a subject within an image, determines a priority part, and extends the subject region from the priority part in the direction of detected parts, allowing for accurate subject selection even when non-detection regions are involved.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple parts are detected for one subject, then subject detection accuracy is improved, but misselection occurs when non-detection parts are present among detected parts
Solution Approach 1:
The patent segments the subject into multiple detection parts (e.g., face, body, limbs) and detects each part independently. By dividing the subject detection into multiple parts, the system can identify individual components while using their spatial relationships to distinguish actual subject parts from unrelated objects, thereby resolving the misselection problem while maintaining detection accuracy
Solution Approach 2:
The patent introduces an intermediary selection mechanism that evaluates the spatial relationships and consistency between multiple detected parts. This intermediary layer acts as a mediator between raw detection results and final subject selection, verifying whether detected parts form a coherent subject structure before confirming the selection, thus preventing misselection caused by non-detection parts
2Ease of operation
If a user performs touch operation on non-detection part, then user input is accepted, but desired subject cannot be selected
Solution Approach 1:
The patent implements a feedback mechanism where the system provides visual or informational feedback to the user about which parts are valid detection targets. When a user performs a touch operation, the system verifies whether the touched region corresponds to a detected subject part and provides feedback accordingly, guiding the user to select valid subjects while maintaining ease of operation
Solution Approach 2:
The patent performs preliminary verification of user input by checking whether the touched region corresponds to a detected subject part before processing the selection. This preliminary action prevents misselection by validating the user's touch input against the detected parts data structure, ensuring that only valid subject parts can be selected while maintaining user-friendly operation
Data Source
AI summary
An image processing device that acquires an image and detects a subject. The image processing device includes at least one processor and a memory holding a program that causes the processor to function as a detection unit configured to detect a plurality of parts related to a same subject within the image, a determination unit configured to determine a priority part from the plurality of parts detected by the detection unit, and an extension unit configured to extend a region of the subject corresponding to a first part from the first part determined as the priority part in a direction of a second part that is different from the first part among the plurality of parts.