(305b) Advances in Computer Vision for Analysis of Coating Appearance and Performance

With digital technologies gaining traction in the materials industry, the traditional workflows for generating and testing paint formulations also continues to evolve. Many critical coating performance attributes are rated visually by researchers to quantify film performance. This approach can be prone to user subjectivity and inconsistency due to limitations of human visual perception and variable assessment conditions. The Dow Coating Materials team has been developing digital workflows to address how visual properties are captured, processed, and referenced for coatings applications. Modern image capture technology coupled with controlled lighting standardizes the collection of visual data into a digital format, with relevant experimental context and metadata tagged in acquired images. Sophisticated processing algorithms leverage these images for robust, quantifiable analysis and produce ratings devoid of user subjectivity. Here we discuss the use of a holistic digital infrastructure using computer vision for select coatings applications. The deployment of such digital tools improves the data quality for impacted test methods and promotes improved access to this visual data through a standardized data format stored in a centralized database.