Thermal signature for solder defect detection using a neural network approach

Sheng-Jen Hsieh; Agapito Calderon

doi:10.1117/12.421049

23 March 2001 Thermal signature for solder defect detection using a neural network approach

Sheng-Jen Hsieh, Agapito Calderon

Proceedings Volume 4360, Thermosense XXIII; (2001) https://doi.org/10.1117/12.421049
Event: Aerospace/Defense Sensing, Simulation, and Controls, 2001, Orlando, FL, United States

Abstract

This paper describes a neural network approach to detecting solder defects on printed circuit boards when using thermal signatures. Solder defects such as open and insufficient solder was investigated. A multilayer neural network with multiple inputs and a single output was utilized. A back- propagation algorithm was utilized within the network. Computer mouse printed circuit boards with known introduced solder defects and amounts of solder were used for experiments. Thermal images were acquired as the boards were powered up. A Visual Basic program was written to retrieve temperature data from an encoded image file format. Afterwards, MATLAB neural network routines were applied to analyze the data. The neural network was able to diagnose solder defects on two of five resistors with 91.1% accuracy, and on three of five resistors with 61.1% accuracy.

Citation Download Citation

Sheng-Jen Hsieh and Agapito Calderon "Thermal signature for solder defect detection using a neural network approach", Proc. SPIE 4360, Thermosense XXIII, (23 March 2001); https://doi.org/10.1117/12.421049

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available