AM62A system-on-chip (SoC) is used to build an end-end application for defect detection in manufacturing. AM62A is a heterogeneous processor equipped with a 2 TOPS Deep Learning Accelerator and up to four Arm® Cortex® A53 processors in addition to various other accelerators for Video and Vision processing. The various compute cores and rich peripheral set make AM62A an ideal option for applications where advanced sensor processing capability is required in real-time. This document describes the complete process of building a defect detection application starting from data collection, deep learning model selection, model training and model deployment. It shows how TI’s EdgeAI Studio tools simplify this process. System level performance analysis of the application, resource utilization and power profiling using TI's tools are presented. Source code and a step-by step guide in TI’s github repository are also available and links are provided for the interested developer at: https://github.com/TexasInstruments/edgeai-gst-apps-defect-detection.
Sitara™ is a trademark of Texas Instruments.
Arm® and Cortex® are registered trademarks of Arm Limited (or its subsidiaries) in the US and/or elsewhere.
All trademarks are the property of their respective owners.
Defect detection is a crucial part of quality assurance in the manufacturing process. This demo uses AM62A to run a vision based artificial intelligence model for defect detection for manufacturing applications. The model tests the produced units as they move on a conveyor belt, to recognize the accepted and the defected units. Figure 1-1 shows a screenshot of the application.
An object tracker is developed for this demo to provide accurate coordinates of the units for sorting and filtering. A live video is displayed on the screen. The units are marked on the screen using green boxes for good (accepted) units while defected units are marked with boxes with different shades of red to distinguish the types of defects. The screen also includes a graphical dashboard showing live statistics about total products, defect percentage, production rate, and a histogram of the types of defect. The object tracker and the graphical dashboard are built using Python. The code base and details of how to run the demo are available on TI marketplace as a Github repo: https://github.com/TexasInstruments/edgeai-gst-apps-defect-detection.
The steps followed to develop this application includes: