GPU-Based Image Defect Detection for High-Throughput Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing manufacturing defect detection systems face bottlenecks due to the reliance on central processing units (CPUs) for image processing, leading to delays and reduced manufacturing throughput, especially in high-speed environments.
Innovation Solution
A system utilizing graphics processing units (GPUs) with a data processing unit (DPU) to process selective image sections directly, bypassing CPU intervention, enabling efficient defect detection and analysis by associating GPUs with specific image processing tasks based on item features or defect types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If CPUs are used for image processing in defect detection systems, then image processing can be performed, but processing speed is slow and manufacturing throughput is reduced
Solution Approach 1:
The patent segments the image processing workload by dividing images into multiple sections and assigning different sections to different GPUs for parallel processing. This segmentation enables simultaneous processing of multiple image sections, dramatically increasing processing speed and maintaining high manufacturing throughput without bottlenecking on CPU performance.
2Productivity
If all images are processed by a single CPU, then centralized control is maintained, but processing capacity is limited and latency increases
Solution Approach 1:
The patent merges multiple GPUs into a unified defect detection system, where each GPU processes specific image sections concurrently. This combination of multiple processing units dramatically increases overall processing capacity and reduces latency compared to a single CPU, while still maintaining coordinated control through the system architecture.
3Speed
If multiple GPUs process different image sections, then processing speed increases, but system complexity increases
Solution Approach 1:
The patent applies local quality by assigning specific image processing tasks to specific GPUs based on their capabilities and the characteristics of different image sections. Each GPU is optimized for processing particular types of image data, which increases processing speed while managing system complexity through specialized rather than generalized processing units.
Data Source
AI summary
Systems and methods herein are for image-based defect detection for items moving through a manufacturing area. A data processing unit (DPU) is in communication with multiple graphics processing units (GPUs) and can receive images captured of an item, as well as receive an indication of image sections of the images associated with a characteristic of a manufacturing process of the item. The DPU can provide the image sections to the GPUs for image processing to determine a value associated with the characteristic of the manufacturing process of the item. The provision of the image sections may be based at least in part on the individual ones of the GPUs being associated with a feature of the item, and the GPUs can perform different image processing based at least in part on different features of the items.