Computer vision method for improved automated image capture and analysis of rapid diagnostic test devices
Abstract:
The disclosed embodiments are generally directed to improving feature detection of rapidly acquired images using camera-enabled mobile devices involving a 2-D decal code, such as a QR code, for improving the reading accuracy of a rapid diagnostic antigen or antibody or enzymatic colorimetric directed test, such as for COVID-19 diagnosis. One primary issue with evaluating a Covid-19 rapid test is detecting and quantifying positive test lines from sampled test strips based on digital images of the test strip. Aspects of the present invention contemplate masking a QR code to improve the sample image resolution and contrast. Other aspects of the present invention contemplate methods and techniques to evaluate a test line on the sample image by enhancing an intensity curve along the test line and control line containing area by way of calculating the instantaneous change in pixel intensity and evaluating the position and intensity of those signals.
Information query
Patent Agency Ranking
0/0