Microscopy School Lesson 11 – Principles of Deconvolution
Whenever a measurement of a physical variable is made, the response function of the instrument impacts the result. Instrument response functions are generally described as filters, which reduce resolution in one way or another. In linear systems, we model the measurement process mathematically as Convolution. If we have knowledge of the instrument response function then we can build computational algorithms to reduce the impact of the instrument and deliver results with greater fidelity. The computational process and algorithms are collectively known as Deconvolution. The approach is applicable across all spheres of measurement science.
In this presentation Dr Mark Browne will focus on optical microscopy and outline concepts of image formation by convolution; microscope point spread and optical transfer functions; and how these can be applied to develop simple and iterative deconvolution algorithms. Dr Browne will present examples and identify common pitfalls to be avoided.
Important parameters in a microscope
Image formation by convolution
Point spread function and optical transfer function