Image Processor Position-Based Attention Object Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing systems struggle to identify attention objects in photographs that are not included in pre-selected sample images, requiring manual user input to specify subjects of interest.
Innovation Solution
An image processor that includes an image input section, position information acquisition section, object recognition section, attention object list management section, and attention object identification section, which compares recognized subjects with pre-photographed attention object data to automatically identify attention objects based on priority and frequency of recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If the system uses pre-selected sample images to identify attention objects, then the identification process is simplified, but subjects not included in pre-selected samples cannot be identified as attention objects
Solution Approach 1:
The system performs preliminary actions by pre-selecting sample images and extracting their subjects before actual photographing. This creates a database of attention object candidates that speeds up subsequent identification processes while maintaining the ability to adapt to new subjects through manual specification and learning mechanisms.
Solution Approach 2:
The system enables self-service by automatically learning and updating attention object candidates from user-specified subjects. When users manually specify attention objects, the system automatically adds them to the candidate pool, eliminating the need for continuous manual configuration and enabling the system to adapt to new subjects autonomously.
2Measurement precision
If the system requires manual user input to specify attention objects, then accuracy in identifying user-interest subjects is improved, but operation complexity and time consumption increase
Solution Approach 1:
The system applies partial action by requiring user input only when necessary - specifically when subjects are not found in the pre-selected sample images or when users want to emphasize particular subjects. For routine cases matching sample images, the system operates automatically without requiring user input, thus maintaining accuracy while reducing operational burden.
Solution Approach 2:
The system implements feedback mechanisms where user specifications of attention objects are fed back into the system to update the attention object candidate database. This feedback loop allows the system to learn from user preferences and improve future identification accuracy automatically, reducing the need for repeated manual inputs.
3Measurement precision
If the system compares recognized subjects with attention object candidates from previously photographed images at the same position, then identification accuracy for frequently photographed subjects is improved, but device complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-processing and storing subject recognition results along with their associated position information from previously photographed images. This creates a structured database that enables efficient position-based queries and comparisons, improving identification accuracy while managing complexity through organized data storage and retrieval mechanisms.
Data Source
AI summary
An image processor has an image input section that receives image data output from an imaging unit for imaging a subject, the image data including the subject, a position information acquisition section that acquires position information indicating a current position, an object recognition section that recognizes the subject in the image data input from the image input section, and an attention object list management section that manages, for each position upon photographing, a first attention object list including first attention object data indicating an attention object candidate selected from among the subjects recognized in the image data that is previously photographed at the position upon photographing.