Image Subject Classification for Overlapping Detection Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle with accurately determining the type of a subject when multiple detection results overlap in a region, often leading to false detections.
Innovation Solution
An image processing apparatus and method that utilizes a convolutional neural network (CNN) to evaluate the certainty of a subject type by comparing detection results using a type estimation CNN, allowing for accurate determination of the main subject even when different types of subjects overlap.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple types of subject detection are performed simultaneously, then the versatility of subject detection is improved, but the reliability of detection results deteriorates due to false detections in overlapping regions
Solution Approach 1:
An evaluation value acquisition unit is introduced as an intermediary component that calculates confidence scores for each detection result. This mediator evaluates the reliability of overlapping detections by comparing evaluation values from different detection types, allowing the system to resolve conflicts and select the most reliable detection result while maintaining multi-type detection capability
2Area of stationary object
If subject detection is performed in overlapping regions, then the coverage of detection is improved, but the measurement precision of subject type deteriorates due to conflicting detection results
Solution Approach 1:
The system implements feedback by acquiring evaluation values for each detection result and using these values to determine the final subject type. The evaluation values provide feedback on the confidence level of each detection, allowing the system to adjust its decision-making process and select the detection result with the highest confidence, thereby maintaining precision even in overlapping regions
Data Source
AI summary
An image processing apparatus acquires an image, detects a plurality of types of subjects included in the image, acquires an evaluation value indicating certainty of a type of a subject for a region of a detected subject; and evaluates a type of a subject for a region of a subject detected. In a case where a region of a different type of subject overlaps with a region of a first subject that is detected, the apparatus acquires the evaluation value and evaluates a type of the first subject based on the evaluation value.