Neural Network Processor Selective Activation for Image Quality and Power
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processing devices face inefficiencies in power consumption and quality deterioration due to increased demand for high-quality images, as existing image processors incur high overhead and are not optimized for selective use of neural network processing.
Innovation Solution
An image processing device that utilizes a neural network processor for specific image processing modes, such as despeckling and denoising, while allowing the main processor to perform other computations efficiently, and selectively deactivating the neural network processor for power conservation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the neural network processor is used for all image processing tasks to improve image quality, then image quality is improved, but power consumption increases
Solution Approach 1:
The patent changes the operational parameters of the neural network processor by selectively activating it only for specific processing modes (e.g., when noise reduction is needed) rather than continuously. The controller determines whether to activate the neural network processor based on image characteristics, thereby optimizing power consumption while maintaining image quality when necessary
Solution Approach 2:
The system dynamically adjusts the processing architecture by switching between using the neural network processor and the main processor based on real-time requirements. The neural network processor is activated only when specific processing tasks require its capabilities, creating a dynamic and adaptive power management strategy
2Productivity
If the neural network processor is activated for all tasks to improve processing capability, then processing capability is improved, but device complexity increases
Solution Approach 1:
The patent segments the processing tasks by function and complexity, assigning specific tasks to the neural network processor while leaving other tasks to the main processor. This functional segmentation allows the system to leverage specialized processing capabilities only when needed, reducing overall system complexity while maintaining high processing capability for specific tasks
3Device complexity
If the main processor handles all image processing to reduce device complexity, then device complexity is reduced, but power consumption increases
Solution Approach 1:
The controller acts as an intermediary that intelligently routes image processing tasks to either the main processor or the neural network processor based on task requirements. This intermediary component enables efficient task distribution, allowing the system to use the power-efficient neural network processor for suitable tasks while maintaining low overall device complexity
Data Source
AI summary
An image processing device includes: an image sensor configured to generate first image data by using a color filter array; and processing circuitry configured to select a processing mode from a plurality of processing modes for the first image data, the selecting being based on information about the first image data; generate second image data by reconstructing the first image data using a neural network processor based on the processing mode; and generate third image data by post-processing the second image data apart from the neural network processor based on the processing mode.