Visual Feedback Loop for Part Profile Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying parts without identification tags are inaccurate, particularly for parts with complex surface geometries, due to limitations in computer vision and laser scanning technologies.
Innovation Solution
A method and system that utilize a visual feedback loop between a display and an image capturing device to refine a part profile picture, allowing for accurate identification of parts by projecting patterned images and adjusting colors to enhance contrast.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional measurement tools are used to measure part surfaces and features, then the measurement process is simple and fast, but the measurement accuracy is poor particularly for complex surface geometries
Solution Approach 1:
The patent introduces a display screen as an intermediary between the camera and the part. The display projects a dynamic pattern (such as a moving grid or contour) that interacts with the part's surface geometry. By analyzing how the pattern deforms or is occluded by the part, the system can accurately capture complex surface features without requiring sophisticated measurement hardware.
Solution Approach 2:
The system employs feedback by continuously monitoring the captured images and adjusting the projected pattern on the display accordingly. The pattern adapts in real-time to enhance contrast at boundaries and features of interest, allowing the system to iteratively improve measurement accuracy for complex geometries.
2Measurement precision
If computer vision methods are used for part identification, then the system is simple and cost-effective, but the identification accuracy is poor due to blurred boundaries from poor lighting or incompatible background coloring
Solution Approach 1:
The patent utilizes color changes by projecting patterns with specific color contrasts onto the part. The display can dynamically adjust the colors of the projected pattern to ensure maximum contrast against the part's surface and background. This allows the image capturing device to clearly distinguish part boundaries and features even in varying lighting conditions.
Solution Approach 2:
The projected pattern on the display is dynamic rather than static. It can move, change intensity, or adjust its spatial distribution based on real-time feedback from the image capturing device. This dynamic adaptation allows the system to optimize boundary detection and overcome issues with poor lighting or incompatible backgrounds.
3Measurement precision
If laser scanners are used for precise part measurement, then the measurement precision is high, but the system becomes expensive and less scalable
Solution Approach 1:
The patent replaces mechanical laser scanning systems with an optical projection and image capture system. Instead of using physical laser beams that require precise mechanical positioning, the system projects optical patterns through a display and uses a stationary or minimally moving camera to capture the interactions. This substitution maintains measurement precision while dramatically reducing system complexity and cost.
Solution Approach 2:
The system creates a visual copy of the measurement process by projecting a digital pattern onto the physical part and capturing the interaction optically. This allows precise measurement information to be obtained through optical copying rather than direct mechanical contact or complex laser scanning, making the system more scalable and cost-effective.
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
This approach enables expedient and highly accurate identification of unmarked parts, overcoming the limitations of previous technologies by improving precision and adaptability to complex geometries.
Implementation Method 1
an image capturing device pointed toward the display... The image capturing device is oriented to capture images of the part with the display screen in background behind the part
Data Source
AI summary
A part is positioned between a display and an image capturing device pointed toward the display. A visual feedback loop is conducted using the display and the image capturing device to refine a part profile picture on the display corresponding to a profile of the part in an image captured by the image capturing device. The visual feedback loop recursively projects on a display screen an input picture containing a pattern region; captures images of the part while the display is displaying the input picture; based on the captured image, refines a picture that more closely corresponds to the profile of the part; and sets the refined image to be a next input image. The feedback loop iteratively refines the picture projected on the display screen until it becomes the part profile picture.