Dual-Circuit Subject Tracking for Low-Power Image Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The use of machine learning for subject tracking in battery-powered devices like image capture apparatuses leads to high power consumption, reducing operable time due to the need for high-performance circuits.
Innovation Solution
Implementing a combination of machine learning (ML)-based and non-ML-based tracking processing, with controlled operation frequencies to balance accuracy and power consumption, using an ML tracking circuit and a non-ML tracking circuit, where the ML tracking circuit operates at a lower frequency than the non-ML tracking circuit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning-based tracking processing is used to improve tracking accuracy, then tracking precision is improved, but power consumption increases
Solution Approach 1:
The patent dynamically adjusts the operation frequency of the ML tracking circuit based on tracking accuracy requirements and power consumption conditions. The control unit monitors tracking accuracy and selectively activates the ML tracking circuit only when necessary, rather than operating continuously at fixed frequency, thereby resolving the contradiction between maintaining high tracking accuracy and reducing power consumption.
Solution Approach 2:
The patent changes the operational parameters of the tracking system by adjusting the operation frequency of the ML tracking circuit. The control unit modulates the frequency at which the ML tracking circuit processes images based on current tracking accuracy needs and power consumption constraints, allowing the system to adapt between high-accuracy mode and low-power mode dynamically.
2Measurement precision
If machine learning-based tracking processing is used to improve tracking accuracy, then tracking precision is improved, but operable time is reduced
Solution Approach 1:
The patent implements periodic operation of the ML tracking circuit rather than continuous operation. The control unit determines specific time periods or intervals when the ML tracking circuit should operate based on tracking accuracy requirements, allowing the non-ML tracking circuit to handle routine tracking and extending battery-operable time while maintaining accuracy when needed.
Solution Approach 2:
The system dynamically adjusts the operational duration and timing of the ML tracking circuit based on real-time tracking accuracy requirements and remaining battery capacity, optimizing the balance between maintaining high tracking precision and extending overall operable time of the device.
3Measurement precision
If high-performance circuit is used to enable machine learning processing, then tracking accuracy is improved, but power consumption increases
Solution Approach 1:
The patent segments the tracking processing into two distinct circuits: a non-ML tracking circuit for routine tracking operations and an ML tracking circuit for high-accuracy tracking when needed. This segmentation allows the high-performance ML circuit to be used only periodically rather than continuously, reducing overall power consumption while maintaining high tracking accuracy when required.
Solution Approach 2:
The control unit acts as an intermediary that manages the operation of both tracking circuits. It determines when to activate the high-performance ML tracking circuit based on tracking accuracy requirements and power consumption conditions, mediating between the two circuits to optimize the balance between tracking accuracy and power consumption.
Data Source
AI summary
An image processing apparatus comprises a first tracking circuit configured to apply machine learning (ML)-based first tracking processing to images and a second tracking circuit configured to apply non-ML-based second tracking processing to the images. The apparatus controls operations of the first and second tracking circuits so that an operation frequency of the first tracking circuit to be lower than an operation frequency of the second tracking circuit.