An iterative non-negative matrix factorization filter(INNMFF) algorithm was developed to be used as a digital filter for deconvolving photon arrival events from highly convolved time traces. Photon counting (PC) is a common method for increasing signal to noise for low intensity measurements. Unfortunately, PC exhibits a limited dynamic range due to pulse pile up which is caused by the convolution of the impulse response function (IRF) and the instantaneous photon arrival event. NNMF is a common method for deconvolving signals of interest from a convoluted time trace, where in some cases the signal can be solved for by matrix inversion and multiplication. Consequently, the process belongs to the class of NP-Complete problems which require exponentially more computational time with increasing data size. Adapting NNMF into an iterative digital filter significantly decreases the computational time required to recover both the photon arrival times and the impulse response function. The INNMFF was applied to triboluminescence time traces consisting of ~87 million data points and completed recovery in ~20 minutes, while conventional NNMF methods were unable to analyze a data trace with 2.5 million data points.<br/>

Non-negative matrix factorization filter for digital deconvolution

256th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.