Automated droplet measurement (ADM): an enhanced video processing software for rapid droplet measurements

Abstract

This paper identifies and addresses the bottlenecks that hamper the currently available software to perform in situ measurement on droplet-based microfluidic. The new and more universal object-based background extraction operation and automated binary threshold value selection make the processing step of our video processing software (ADM) fully automated. The ADM software, which is based on OpenCV image processing library, is made to perform measurements with high processing speed using efficient code. As the processing speed is higher than the data transfer speed from the video camera to permanent storage of computer, we integrate the camera software development kit (SDK) with ADM. The integration allows simultaneous operations of the video transfer/streaming and the video processing. As a result, the total time for droplet measurement using the new process flow with the integrated program is shortened significantly. ADM is also validated by comparing with both manual analysis and DMV software. ADM will be publicly released as a free tool. The software can also be used on a video file or files without the integration with the camera SDK.