Computer Vision Processing Circuitry with Adaptive Power Domains
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices with transparent displays face challenges in efficiently managing power consumption and image processing tasks, particularly when capturing and displaying images for both environmental awareness and user interaction, without compromising performance.
Innovation Solution
The device segregates image signal processing into a first power domain for computer vision processing and a second power domain for high-quality image processing, selectively activating the high-quality pipeline only when necessary, and employs a multiclient scheduler to manage image requests and reuse captured images to optimize power usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-quality image processing pipeline is continuously activated, then image display quality is improved, but power consumption increases
Solution Approach 1:
The system dynamically switches between high-quality and low-power processing modes based on real-time display needs. The back-end image signal processing pipeline is selectively deactivated when displays are not outputting content, and activated only when high-quality image processing is required, making the processing quality adaptive rather than static.
Solution Approach 2:
The system employs periodic activation of the high-quality processing pipeline rather than continuous operation. The back-end pipeline is activated in periodic intervals when display content is output and deactivated during intervals when no content is displayed, reducing overall power consumption while maintaining quality when needed.
2Productivity
If multiple image processing clients run simultaneously, then processing capability is improved, but resource management complexity increases
Solution Approach 1:
The patent introduces a multiclient scheduler as an intermediary component that manages image requests from multiple clients. The scheduler receives requests, reorders them based on deadlines and priority levels, and coordinates resource allocation, thereby simplifying the management complexity while enabling multiple clients to run simultaneously with improved processing capability.
3Ease of operation
If image requests are processed in strict FIFO order, then fairness is improved, but timing requirements may not be met
Solution Approach 1:
The scheduling system dynamically reorders image requests in the queue based on their specific requirements. Instead of a static FIFO approach, the scheduler adjusts the processing order considering deadlines, timing requirements, and priority levels, allowing critical requests to be processed earlier while maintaining fairness through priority-based management.
Solution Approach 2:
The system changes the scheduling parameter from simple FIFO ordering to a multi-parameter sorting mechanism that considers deadlines, timing requirements, and priority levels. This parameter change enables the system to meet timing requirements while maintaining fairness through priority-based request handling.
Data Source
AI summary
An electronic device is provided that can include sensors for capturing an image, computer vision processing (CVP) circuitry that receives the captured image and having subsystems operating in a first power domain, a back-end image signal processing pipeline operating in a second power domain different than the first power domain, and optionally displays configured to receive content for display from the back-end image signal processing pipeline. A multiclient scheduler in the CVP circuitry can receive and reorder image requests received from a plurality of client processors in the electronic device. The multiclient scheduler can query an image server on the electronic device to determine whether image requests can be satisfied by an existing image currently stored on the image server.