Image sampling


In microscopy, image quality is not only limited by optics, but also by how finely the image is sampled. Even a perfect microscope can produce misleading data if the sampling is inappropriate. Under-sampling throws away biophysical information that can never be recovered, while over-sampling creates the illusion of higher resolution at the cost of time, light dose, and storage. Understanding image sampling means knowing how much information your experiment actually contains — and how much you really need to record.


Prerequisites

Before starting this lesson, you should be familiar with:

Learning Objectives

After completing this lesson, learners should be able to:

  • Explain how spatial and axial sampling discretize a continuous optical image.

  • Identify under-sampling artifacts such as aliasing and loss of structural information.

  • Recognize over-sampling and its practical costs (data volume, phototoxicity, acquisition time, processing time).

  • Understand that the biological question guides the required sampling (task-based sampling).

Concept map

graph TD A[Continuous Optical Image] -->|Sampling| B[Discrete Pixels & Voxels] B --> C{Sampling Regime} C -->|Under-sampling| D[Aliasing & Information Loss] C -->|Task-based sampling| I[Sufficient Representation] C -->|Optimal sampling| E[Faithful Representation] C -->|Over-sampling| F[Redundant Data] F --> G[Longer Acquisition / Photobleaching] D --> H[Biased Measurements]

Figure


(Left) The pyramid of frustration. Imaging is a compromise where a user has to decide where the available photon budget goes. (Right) DNA acquired at different spatial sampling; scale bar 10 micrometer. Top Left: Intranuclear structures can be investigated. Middle: Intracellular structures are not visible but the number of nuclei could still be measured. Right: Nuclei start to blur so much rendering cell counting challenging. The lower panel is a zoom in.



Activities

Downsample an image

A good way to explore what one can still see at different spatial samplings is to acquire an image with very fine sampling and then downsample it in a software. This can inform you how you would then acquire more images with optimal sampling of the microscope.

The two example images show the Z-maximal projection of several cells at a pixel-size of 35 nm (Zeiss Airy scan).


Show activity for:  

ImageJ Macro

run("Close All")

// Open the image
//open("https://github.com/NEUBIAS/training-resources/raw/master/image_data/xy_16bit__DNA_sampling_35nm.tif")
open("https://github.com/NEUBIAS/training-resources/raw/master/image_data/xy_16bit__MT_sampling_35nm.tif")
rename("scaling_1");
//Down scale with a factor of 2
selectImage("scaling_1");
run("Scale...", "x=0.5 y=0.5 interpolation=Bilinear average create title=scaling_2.tif");

//Down scale with a factor of 4
selectImage("scaling_1"); // Make sure you always start from the original image
factor = 4
downsample = 1/factor
run("Scale...", "x="+downsample+" y="+downsample+" interpolation=Bilinear average create title=scaling_"+factor+".tif");

//Down scale with a factor of 16
selectImage("scaling_1");
factor = 16
downsample = 1/factor
run("Scale...", "x="+downsample+" y="+downsample+" interpolation=Bilinear average create title=scaling_"+factor+".tif");






Assessment

Fill in the blanks

  1. Over-sampling does not improve true optical resolution, but it increases _____, _____, and _____.

Solution

  1. under; lost
  2. data volume; acquisition time; photobleaching




Follow-up material

Recommended follow-up modules:

Learn more: