Robot Vision System Asynchronous Parallel Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic vision systems are limited in their ability to recognize and track objects in 3D environments with six degrees of freedom, relying on static image processing and fixed algorithms, which hinders their flexibility and efficiency in dynamic manufacturing settings.
Innovation Solution
A machine vision system with a controller and image acquisition sensor that captures 3D images, featuring a focus loop control routine for dynamic object tracking and a background loop for environmental monitoring, allowing for asynchronous parallel execution and efficient distribution of computing resources based on priority.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a single-threaded vision system executes a sequence of algorithms, then the system structure is simple, but the processing speed and real-time tracking capability are insufficient
Solution Approach 1:
The vision system is segmented into multiple independent threads: a first thread dedicated to real-time tracking of the focus object, and a second thread for monitoring background objects. This segmentation allows parallel execution of algorithms, improving processing speed while maintaining manageable system complexity through modular design.
Solution Approach 2:
The system dynamically allocates computing resources based on priority, with the focus loop receiving higher priority for real-time tracking. The thread scheduling and resource distribution are dynamic rather than static, allowing the system to adapt to varying computational demands and maintain high processing speed for critical operations.
2Adaptability or versatility
If vision systems use fixed processing steps, then the system is stable and easy to control, but the adaptability to new vision algorithms and environments is poor
Solution Approach 1:
The vision system is designed with universal, modular components that can execute different vision algorithms. The controller is capable of running multiple types of vision processing routines, allowing the system to adapt to new algorithms and environmental conditions while maintaining a consistent architectural framework.
Solution Approach 2:
The system employs dynamic resource allocation and priority-based scheduling that can adapt to different algorithmic requirements. Computing resources are distributed based on the specific needs of each vision algorithm being executed, enabling flexibility in handling diverse algorithms without requiring complete system redesign.
3Measurement precision
If the vision system focuses only on tracking a single object, then the tracking accuracy is high, but the ability to recognize other objects in the workspace is lost
Solution Approach 1:
The vision system segments the workspace into a focus region and background region, with dedicated processing threads for each. The first thread maintains high tracking accuracy for the focus object, while the second thread monitors and recognizes background objects, thereby preserving both tracking precision and overall object recognition capability.
Solution Approach 2:
The background loop acts as an intermediary that bridges the focus tracking and overall scene understanding. It monitors objects other than the focus object and can transfer relevant information to the focus loop, enabling the system to maintain accurate focus tracking while remaining aware of and capable of recognizing other objects in the workspace.
4Productivity
If computing resources are distributed based on priority with parallel threads, then the real-time tracking performance is improved, but the system complexity increases
Solution Approach 1:
The system segments computing resources into distinct threads with clearly defined responsibilities: one thread for focus object tracking and another for background monitoring. This segmentation simplifies thread management by creating well-defined, independent processing units rather than requiring complex inter-thread coordination for all operations.
Solution Approach 2:
Each thread is designed to be relatively self-sufficient, with the focus loop independently handling real-time tracking of the focus object and the background loop independently monitoring other objects. This self-service approach reduces the need for complex inter-thread communication and coordination, thereby managing system complexity while maintaining high productivity.
Data Source
AI summary
A machine vision system for a controllable robotic device proximal to a workspace includes an image acquisition sensor arranged to periodically capture vision signal inputs each including an image of a field of view including the workspace. A controller operatively couples to the robotic device and includes a non-transitory memory component including an executable vision perception routine. The vision perception routine includes a focus loop control routine operative to dynamically track a focus object in the workspace and a background loop control routine operative to monitor a background of the workspace. The focus loop control routine executes simultaneously asynchronously in parallel with the background loop control routine to determine a combined resultant including the focus object and the background based upon the periodically captured vision signal inputs. The controller is operative to control the robotic device to manipulate the focus object based upon the focus loop control routine.