Vision Sensor with Dedicated Microprocessor for Low Power Motion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging systems in electronic devices face challenges in detecting and characterizing object motion while maintaining low power consumption, especially in low power modes, and require improved techniques for recognizing features like faces and gestures at high frame rates.
Innovation Solution
An imaging system with a dedicated microprocessor and a lens assembly that includes a pixel array and aspherical lenses, capable of performing computer vision computations in-pixel, allowing for low power motion detection and object characterization, even in power-saving modes, with a small form factor and high frame rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a conventional imaging system with full processor is used for motion detection, then detection accuracy and object recognition capability are improved, but power consumption increases significantly
Solution Approach 1:
The system divides the imaging functionality into two segments: a low-power vision sensor with dedicated microprocessor for basic motion detection and object recognition, and a full processor that remains in sleep mode. This segmentation allows the device to perform recognition tasks using only the low-power segment, resolving the contradiction between accuracy and power consumption.
Solution Approach 2:
A dedicated microprocessor acts as an intermediary between the vision sensor and the full processor. It performs computer vision computations locally and only activates the full processor when necessary, enabling accurate object recognition while maintaining low power consumption during normal operation.
2Use of energy by moving object
If the device enters power-saving mode to reduce power consumption, then energy usage decreases, but motion detection and object recognition functionality is lost
Solution Approach 1:
The system segments processing capabilities so that motion detection and basic recognition can be performed by the low-power vision sensor independently, allowing the device to maintain detection functionality even when the main processor is in sleep mode.
Solution Approach 2:
The vision sensor with its dedicated microprocessor serves itself by performing computer vision computations locally without requiring the main processor. This self-service capability enables continuous motion detection and object recognition in power-saving mode.
3Measurement precision
If a high frame rate is used for detecting fast-moving objects, then detection accuracy is improved, but power consumption and processing load increase
Solution Approach 1:
The system extracts only the essential computation tasks (computer vision computations for motion detection) from the main processor and assigns them to the dedicated microprocessor. This extraction enables high frame rate processing without overloading the power-constrained system.
Solution Approach 2:
The dedicated microprocessor is optimized with specific hardware parameters (such as 1.2 GHz clock speed and specialized vision processing units) that enable high frame rate operation at lower power consumption compared to a general-purpose processor running at higher speeds.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables efficient detection and recognition of object motion and characteristics at low power consumption, supporting always-on functionality in electronic devices, including smartphones and tablets, with minimal power draw and high frame rates, even in sleep modes.
Implementation Method 1
a lens assembly optically coupled with the pixel array
Data Source
AI summary
A vision sensor includes a sensor assembly and a dedicated microprocessor. The sensor assembly includes a pixel array and a lens assembly that is optically coupled with the pixel array. The lens assembly has an F#<2, a total track length less than 4 mm, and a field of view of at least +/â20 degrees. The dedicated microprocessor is configured to perform computer vision processing computations based on image data received from the sensor assembly and includes an interface for a second processor.