Deformable Sensor Calibration Using Image Feedback and Pressure Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deformable sensors with short working distances face challenges in accurate object and pose detection due to errors caused by improper inflation or shape of the deformable membrane.
Innovation Solution
A method and system for calibrating deformable sensors using external image sensors to capture image data, compare it to a metric, and adjust the pressure within the enclosure until the image data satisfies the metric, ensuring accurate size and shape of the deformable membrane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If the deformable membrane is inflated to a larger size, then the sensor can detect objects more easily, but the working distance becomes too large causing detection errors
Solution Approach 1:
The patent applies preliminary action by performing calibration of the deformable membrane size and shape before actual object detection. The system captures images of the deformable sensor, compares them to reference images, and adjusts inflation pressure beforehand to ensure the membrane is at the correct size and shape for accurate detection, thus preventing detection errors before they occur
Solution Approach 2:
The patent implements feedback by using an image sensor to continuously monitor the size and shape of the deformable membrane, comparing the captured images to reference images, and adjusting the inflation pressure based on the comparison results. This closed-loop feedback system ensures the membrane maintains the optimal size and shape for accurate object and pose detection
2Volume of moving object
If the deformable membrane is inflated to a smaller size, then the working distance is reduced for better detection, but the sensor may not detect objects effectively
Solution Approach 1:
The system performs preliminary calibration to determine the optimal inflation pressure that achieves the correct balance between size and detection capability. By capturing images and comparing them to reference images before operation, the system ensures the membrane is inflated to the precise size needed for reliable object detection without being too small or too large
3Ease of operation
If the deformable sensor is not properly calibrated, then the system is simpler to operate, but errors are generated in detection
Solution Approach 1:
The patent applies self-service by implementing an automated calibration system that performs its own calibration without requiring manual intervention. The system automatically captures images of the deformable sensor, compares them to reference images, determines if calibration is needed, and adjusts the inflation pressure autonomously, making the complex calibration process transparent to the user while maintaining high detection accuracy
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
The calibration method ensures accurate object and pose detection by maintaining the deformable sensor within a threshold size and shape, reducing errors and improving the reliability of deformable sensors in robotic applications.
Implementation Method 1
capturing image data of the deformable sensor using an external image sensor
Data Source
AI summary
Systems and methods for calibrating deformable sensors are disclosed. In one embodiment, a method of calibrating a deformable sensor includes capturing image data of the deformable sensor using an external image sensor, wherein the deformable sensor comprises a deformable membrane defining an enclosure that is configured to be filled with a medium. The method further includes comparing the image data of the deformable sensor to a metric. When the image data does not satisfy the metric, the method includes adjusting a pressure within the enclosure.