Camera-Guided Component Picking Using Target and Alert Areas
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing assembly systems require significant computational resources to detect the position of a user's hand, which can be inefficient and resource-intensive.
Innovation Solution
A method that captures a series of images and detects changes in specific areas of the images to classify user actions, using assembly sequence information to guide the user in picking components, with minimal computational requirements by focusing on area-based classification and image data comparison.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If hand position detection algorithms are used to track user interactions, then user guidance accuracy is improved, but computational resource consumption increases
Solution Approach 1:
The storage arrangement is divided into multiple storage units, each monitored independently. The camera captures images and the system segments the visual field to identify changes in specific storage units rather than processing the entire scene continuously, reducing computational load while maintaining detection accuracy for relevant areas.
Solution Approach 2:
Instead of continuously tracking hand position throughout the entire workspace, the system performs partial detection by only analyzing specific storage units where component changes are expected. The system captures images at intervals and only processes areas where changes are anticipated, reducing overall computational requirements while maintaining effective monitoring.
2Measurement precision
If continuous image processing is performed to monitor assembly progress, then assembly guidance accuracy is improved, but processing time increases
Solution Approach 1:
The camera captures images of the storage arrangement at periodic intervals rather than continuously. The control device compares successive images to detect changes in storage units, providing updated assembly guidance only when changes are detected. This periodic monitoring approach maintains accurate tracking of assembly progress while significantly reducing processing time and computational overhead.
Solution Approach 2:
The system uses the stored initial image as a reference baseline and automatically compares subsequent captured images against this baseline to detect changes. This self-referential comparison method eliminates the need for continuous external intervention or complex real-time processing, allowing the system to autonomously monitor assembly progress with minimal processing requirements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method (120) for assisting a user in assembling components from one or more storage units (106) of a storage arrangement (104) into a product (102) in accordance with assembly instructions, the method (120) including: Providing (S10) an assembly sequence information indicative of one or more picking actions to be performed by a user; and Performing (S12) for each picking action: Capturing (S14) a sequence of images (122) of the storage arrangement (104); Determining (S20) one or more storage units (106) of the storage arrangement (104) as target storage units (124); Determining (S26) within each image (122) one or more target areas (126), such that a view on each target storage unit (124) corresponds to one target area (126); Determining (S28) within each image (122) one or more alert areas (130), such that each alert area (130) is located in the image within a view on the storage arrangement (104) outside of the one or more target areas (126); Detecting (S52) an alert event if an image data of any alert area (130) of a young image (122) differs from an image data of the same alert area (130) of an old image (122); Detecting (S50) a success event if an image data of any target area (126) of the young image (122) differs from an image data of the same target area (126) of an old image (122); and Outputting (S54) a message to a user interface, wherein the message is indicative of the alert event if the alert event is detected, and wherein the message is indicative of the success event if the success event is detected.