Image Subject Detection Scheduling for High-Frequency Live View Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle to detect and track specific subjects with high frequency, leading to inefficiencies in updating detection frames and handling subject changes during live view image display.
Innovation Solution
A processing apparatus and method that incorporates both subject detection and tracking processes, allowing for repeated first and second detection processing at specific intervals, enhancing the frequency of subject detection and enabling high-frequency updates of detection frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subject detection processing is performed at every frame, then detection frequency is improved, but processing time and computational load increase
Solution Approach 1:
The patent implements periodic subject detection processing where detection is performed at specific intervals (every N frames) rather than continuously at every frame. This periodic approach maintains adequate detection frequency for tracking purposes while significantly reducing the overall computational load and processing time required
Solution Approach 2:
The system performs preliminary subject detection at lower frequency to establish initial tracking parameters, then uses these preliminary results to guide subsequent higher-frequency detection only when necessary. This preliminary action reduces unnecessary processing by preparing the system in advance for when detailed detection is needed
2Loss of time
If subject detection processing is performed at low frequency, then processing time is reduced, but detection frequency and tracking accuracy deteriorate
Solution Approach 1:
The patent implements dynamic adjustment of detection frequency based on subject motion characteristics. When the subject is detected to be moving rapidly or changing position significantly, the system automatically increases detection frequency to maintain tracking accuracy. When the subject is stationary or moving slowly, detection frequency is reduced to save processing time
Solution Approach 2:
The system continuously monitors tracking quality and subject motion patterns, using this feedback to adjust detection frequency in real-time. When tracking accuracy deteriorates or subject motion increases, feedback triggers increased detection frequency. When tracking is stable, feedback reduces detection frequency to optimize processing efficiency
3Reliability
If detection processing is performed frequently, then subject tracking accuracy is improved, but device complexity and computational resources increase
Solution Approach 1:
The patent divides subject detection processing into multiple stages or segments: initial detection phase, tracking phase, and verification phase. Each segment performs specific detection tasks with appropriate frequency and complexity levels, rather than performing full detection processing at every frame. This segmentation maintains tracking accuracy while reducing overall computational complexity
Solution Approach 2:
The system dynamically changes detection parameters such as search window size, detection threshold, and processing intensity based on current tracking conditions. When tracking is stable, parameters are adjusted to reduce processing complexity. When tracking accuracy needs improvement or subject motion increases, parameters are changed to enhance detection performance
Data Source
AI summary
A processing apparatus includes a processor and a memory. The processor performs first detection processing of detecting a first area of a target to be detected from image data acquired from an imaging element and performs second detection processing of detecting a second area based on the first area from image data acquired from the imaging element. The processor further performs first processing using the first area, and performs second processing using the second area. The processor repeatedly performs the first detection processing, and performs the second detection processing in each period from the detecting of the first area started at a first timing to the detecting of the first area subsequently started after the first timing.