Optimising Light Source Positioning for Even and Flux-Efficient Illumination

Designing imaging systems is a challenge faced by researchers in many fields including fluorescent imaging in life sciences and optical set-ups for automated capture of experimental data.

The position of light sources within an imaging system has significant consequences for the downstream usability of the data it generates. Non-uniform illumination can contribute to low quality (or in some cases unusable) images, particularly so when illumination variation approaches or exceeds the sensitivity range of the capture device/camera. Similarly, low flux efficiency (i.e. ratio of flux through the imaging plane divided by the total flux from the source) will negatively affect the image acquisition and subsequent analysis. Often flux efficiency is sacrificed for illumination uniformity i.e. choosing to deliver less light to the target in order to keep the variation of light intensity across the target remains low. Furthermore, the large number of possible positional configurations of a light source within an imaging system precludes manual optimisation.

To tackle this issue, we offer a software for modelling the illumination profile for a given light source. The code can be easily adjusted to model a variety of positional configurations and rapidly calculates results for many thousands of variable combinations. We envisage the exploitation of this software by researchers as well as those in the early stages of prototyping. For example, in the university environment where resources and time are limited.

Furthermore, we demonstrate an approach by which a user can reduce the amount of possible multi-variable combinations down to the most viable options. This is performed using a modified convex hull approach in two-dimensions (optimising for two figures of merit, i.e. the total flux and the illumination variation). In principle, this method can be extended to n-dimensional space to include additional figures of merit for optimisation.