Baseline Correction

Vespucci currently supports three baseline correction methods, and we intend to support more in the future.

Performing Baseline Correction

Supported Methods

Median Filter

The median filter baseline method operates under the assumption that more points exist in the baseline than exist in peaks. The baseline is assumed to be a heavily-smoothed version of the spectrum. The width of the filter window and the number of times to perform smoothing can be selected.

Improved Modified Polynomial Fitting (Vancouver Raman Algorithm)

The Vancouver Raman Algorithm is a modified least-squares polynomial fitting method, originally published here. Our version is a direct implementation from the article without reference to the original code. Because of this, we can't guarantee that it is exactly the same. In fact, we can't really even get the original code to run on our systems to compare.

Rolling Ball

The rolling ball baseline method, originally described here. Vespucci's implementation is a line-for-line translation of this function from the original R.