Data Fusion For Augmented Electronic Counterfeit Detection EfficacyAuthors: Bill Cardoso, Ph.D.
Company: Creative Electron, Inc.
Date Published: 10/13/2013 Conference: SMTA International
The independent use of radiological and optical microscopy for the detection of counterfeit electronic components is a well-established practice in the industry. In this paper we will present the results in fusing radiological and visual data from a set of thousands of components inspected. The scope of this work focuses on two levels of fusion: feature and decision-level.
Feature-level fusion involves the extraction of representative features from sensor data. In feature-level fusion, features from both radiological and optical systems are extracted from multiple observations, and combined into a single concatenated feature vector that is input to pattern recognition approaches based on neural networks.
Decision level fusion involves fusion of sensor information, after each sensor has made a preliminary determination of an entity’s location, attributes, and identity. Examples of features used as entities include pin one location, lead-frame placement, and pin alignment. The decision level fusion engine was fueled with Bayesian inference for final voting determination.
data fusion, counterfeit components, radiography, x-ray inspection, visual inspection.
Members download articles for free:
Not a member yet?
What else do you get when you join SMTA? Read about all of the benefits that go along with membership.
Notice: Sharing of articles is restricted to just your immediate work group. Downloaded papers should not be stored on an external network or shared on the internet.