No beads valid because too close to each other

[rdemets]

Is there a way to reduce the window size on the crop of each detected beads ?
I found an option to crop a x time bigger field, but it doesnt seem to change anything.
I have some images were it looked ok IMO, but there are rejected by the software.

Is there also a way to get an average PSF from all the detected beads ?

Thanks for your help.

[julien-cau]

Hi,

Apologies for the delay; I haven’t been spending much time on GitHub lately.
If the goal is to have a smaller crop box, it’s possible as long as the background annulus geometry is also adjusted. For cropping, the algorithm considers the largest value between the crop factor multiplied by the bead size and the outer edge of the annulus. For example, if you have a bead with a diameter of 0.2 µm, a distance of 0.5 µm between the bead's outer edge and the annulus's inner edge, and an annulus thickness of 1 µm, the minimal box size to fit both the annulus and the bead would be calculated as:
0.2 + 2 × 0.5 + 2 × 1 = 3.2 µm
The algorithm then adds 10% to this value, resulting in a box size of 3.52 µm.
If you set a crop factor of 10, the box size based on the crop factor would be:
0.2 × 10 = 2 µm
However, the algorithm will use the larger of the two values, which in this case is 3.52 µm.

Regarding your comment about averaging beads, I’d be happy to explore this further. However, I don’t currently have much time to investigate averaging methods. Feel free to suggest any approaches you have in mind. My initial thought was to use PALM/STORM algorithms.

Best regards,