(70a) Development of Advanced Image Analysis Algorithms for Quantitative Evaluation of Exterior Coating Degradation | AIChE

(70a) Development of Advanced Image Analysis Algorithms for Quantitative Evaluation of Exterior Coating Degradation

Authors 

Linsen, M., Dow
Long, S., Dow
Kim, S. H., Dow
Hu, Y., University of Wisconsin-Madison
Coating durability, an ability to maintain performance for the intended service period, is an important performance metric for exterior coatings. Traditionally, performance attributes related to durability have been rated by a researcher using a set of guidelines, a method prone to subjectivity. To overcome this challenge, the Dow Coating Materials team has developed a novel workflow to allow quantitative analysis of coating degradation. With the use of new digital image capture technology under a controlled setting and advanced image analysis algorithms, this work provides a robust and quantifiable assessment of coating performance over time. This talk will focus on several uses cases to illustrate the holistic workflow such as: reliably capturing images of exterior coatings, data pre-processing, development of image analysis algorithm, and algorithm deployment. We will also demonstrate the use of both traditional statistical methods and deep learning methods and illustrate how these techniques can be leveraged in different scenarios.