Measure Lengths, Diameters and Areas on Microscope Images
Open a micrograph in your browser, set the scale from its scale bar — or let the file set it — and click to measure. Cells, particles, pores, fibers and layers read in µm, nm or Å, with nothing to install.
Olympus cellSens and JEOL SEM images open already calibrated
Any other micrograph: two clicks on its scale bar
Units from picometers to millimeters, areas in µm² and nm²
Used as a guest, the image never leaves your device
A microscope image has no size until something tells it one. That something is one of three things: a pixel size the instrument wrote into the file, the scale bar burned into the picture, or a scale you add at the microscope before you shoot. Amplia uses whichever you have, then turns every click into a real length.
It is built for the quick, manual part of image analysis: a cell's length, a particle's diameter, a coating's thickness, the area of one grain. For thresholding, automatic counting or intensity, a dedicated analysis suite is still the right tool — more on that below.
Three ways a micrograph gets its scale
Pick the first route that applies to your image. All three end in the same place: a calibrated picture where every tool reads in real units.
1
Open the image
Drop in a JPEG, PNG or TIFF exported from your microscope software. As a guest it is opened in your browser and not uploaded.
2
Set the scale
If the file states its pixel size, this is already done. Otherwise choose the Calibration tool, click both ends of the scale bar, and type its value and unit.
3
Measure
Line for a length, Circle for a diameter, Polygon for an area, Angle for an angle. Each shape gets a numbered label — L1, C1, S1, A1 — and its value.
4
Check one known feature
Measure something whose size you already know before you trust the rest.
Route 1
The file already knows
Some instruments write the size of one pixel into the image file. Amplia reads that record from Olympus cellSens images and JEOL SEM images. Open one and a message says "Calibrated from image metadata", with the pixel size next to it — 16.7 nm/px, for example. No points to place; the Line tool is ready.
The value is applied, not verified. Amplia reads a record another program wrote and cannot check that it belongs to this picture. Measure one known feature if the number matters. You can edit the pixel size, or replace it with your own two-point calibration, at any time.
Every other micrograph with a scale bar takes two clicks: one on each end of the bar, then its value and unit — 50 and µm. From that moment every shape reads in real units, and changing the value later re-scales everything already drawn.
With a free account there is a shortcut: create a project from the image and Amplia looks for the scale bar itself. When it finds one it places the two points on the bar's ends, reads the label, and tells you — "Calibrated from the scale bar" — so you can check it. It reads metric labels; a label it misreads is applied as read, and an image with no readable bar simply opens uncalibrated.
A phone held to the eyepiece, a camera without calibrated software, a frame exported with the overlay switched off: the image itself holds nothing to calibrate on. Fix that at the microscope, not afterwards.
Switch the scale bar on in your acquisition software and export again. This is the cheap fix when the software is calibrated.
If the software was never calibrated, calibrate it there with a stage micrometer — the glass slide with an engraved ruler, usually 1 mm in 100 divisions of 10 µm — once per objective. After that the bar it draws can be trusted.
Shooting through the eyepiece with a phone? Use an eyepiece with a reticle. Its scale appears in every photo, and a stage micrometer tells you what one division is worth at each objective. Calibrate on as many reticle divisions as fit.
Why "×400" is not a scale
Magnification describes what your eye sees through the eyepiece, or a print of one particular size. A digital image is shown at whatever size the screen, the slide deck or the journal chooses, so "×400" stops meaning anything the moment the file leaves the microscope. A scale bar is drawn into the pixels and resizes with them. Trust the bar.
The scale bar is your whole measurement — place it carefully
Every value on the page inherits the two clicks you made on the bar. A sloppy calibration is not one bad number; it is the same percentage error in all of them.
Where does a bar end?
A plain filled bar ends at its outer edges. A bar with end ticks (an I-beam) is ambiguous — the stated length usually runs between the centers of the ticks, sometimes between their outer edges, and the difference is the tick's own thickness. On a project calibrated from the scale bar, the Scale-bar adjustment dialog shows both ends magnified, with a Center ↔ Edge slider that moves both points together. If you don't know which your software draws, calibrate one image both ways and measure something known.
Zoom in before you click
On a desktop, zoom with the wheel until the bar's end fills a good part of the screen. On a phone or tablet, press and hold instead of tapping: a magnified strip opens at the top of the canvas, and a slow finger moves the point more slowly than the finger for fine positioning.
Use the whole bar, and prefer a long one
The error of a one-pixel misclick shrinks with the bar's length in pixels:
How a one-pixel error in placing the calibration points carries into every measurement, by scale bar length
Scale bar length in the image
Error from a 1 px misclick
50 px
2 %
100 px
1 %
200 px
0.5 %
500 px
0.2 %
1000 px
0.1 %
If the bar in your export is short, export at full resolution rather than a downsized preview — the bar gets longer in pixels and the same misclick costs less.
Sanity-check the result
After calibrating, Info → Scale shows what one pixel is worth. Compare it with what your setup predicts:
pixel size ≈ camera pixel pitch × binning ÷ (objective magnification × camera adapter magnification)
A camera with 3.45 µm pixels behind a 20× objective and a 1× adapter gives about 0.17 µm per pixel. If Amplia reports 0.34, the image was binned or downsampled — or the bar was misread.
The tools are the same ones that measure a floor plan. What changes is what you point them at.
Cells and nuclei
Line across the long and short axis for length and width. For a round cell, the Circle tool takes the two ends of a diameter and returns diameter (d), circumference (C) and area (A) together. For an irregular outline, click around it with the Polygon tool — or use the magic wand and let it trace the edge.
One Circle per particle: two clicks on opposite edges. Each gets its own label (C1, C2, …), and the PDF report lists every diameter, circumference and area. This is hand measurement — right for a few dozen particles you want to look at one by one, wrong for a thousand.
Width: a Line straight across, perpendicular to the fiber's axis — a slanted line reads too long. Length along a bent fiber: the Polyline tool, clicking along the bends, gives each segment and the total. A smooth curve through the same points is the Curve tool PRO.
Thickness is a Line from one interface to the other, perpendicular to the layer. Take it at several places — coatings are rarely uniform — and add a comment to each line saying where it was taken. The comments travel into the report.
The area of one region: Polygon by hand, or one click with the magic wand, which outlines the object under the pointer and reports its area. The wand follows the outer edge only — a hole inside the region is not subtracted — so treat its outline as a suggestion and adjust the corners. For counting by eye, switch on the grid and give it a step in real units: squares 10 µm on a side, for instance. The grid is a visual aid and measures nothing itself.
Three clicks: one side, the vertex, the other side. The Angle tool needs no calibration at all, so it works on an image with no scale bar. A polygon can show the interior angle at each corner too, once the image is calibrated.
01The file states its pixel size: 16.7 nm/px. Nothing to calibrate.
02The frame is 1,280 px wide, so the field of view is 1,280 × 16.7 nm ≈ 21.4 µm.
03A Circle across one particle spans 150 px: diameter 2.51 µm.
04From that diameter: circumference 7.87 µm, area 4.93 µm².
Scale from the bar
A cell on a brightfield image
01The scale bar says 50 µm and is 290 px long: one pixel is 50 ÷ 290 ≈ 0.172 µm.
02That agrees with the setup — 3.45 µm camera pixels behind a 20× objective predict 0.1725 µm.
03A Line along the cell's long axis spans 116 px: 20.0 µm.
04A Polygon around the cell encloses 4,200 px²: ≈ 125 µm².
You never do this arithmetic yourself. It is here so you can see that there is nothing hidden in it: a measurement is a pixel count times the scale, and the scale is only as good as the bar or the record it came from.
Micrometers, nanometers, angstroms
Calibrate in the unit on your scale bar. Amplia's metric family runs from picometers to kilometers, and areas follow the same family.
Length units available for microscope images in Amplia, from millimeters to picometers
Unit
Symbol
In meters
In the next unit up
About this big
millimeter
mm
10⁻³ m
—
a grain of coarse sand; a stage micrometer's full ruler
micrometer
µm
10⁻⁶ m
1,000 µm = 1 mm
a red blood cell is 7–8 µm; a human hair 50–100 µm
nanometer
nm
10⁻⁹ m
1,000 nm = 1 µm
a virus is tens to a few hundred nm; a DNA helix is 2 nm wide
angstrom
Å
10⁻¹⁰ m
10 Å = 1 nm
the spacing between atoms in a crystal: 1–5 Å
picometer
pm
10⁻¹² m
100 pm = 1 Å
a carbon–carbon bond is 154 pm
The readable unit is picked for you
With automatic unit conversion on, a 2,505 nm diameter reads 2.51 µm and a 0.5 nm spacing reads 5 Å. Switch it off in Settings to pin every value to the unit you calibrated in — useful when a whole table must be in nanometers.
Areas come out in µm², nm², Ų or pm² on the same rule, and rounding is yours to set: automatic, or a fixed zero to three decimals, separately for length, area and angle. Rounding changes what is shown, never what was measured.
A micrograph is a flat picture of something that is not always flat. The tool measures the picture exactly; whether the picture matches the specimen is a question about the image.
Works well
Flat specimens in one focal plane: thin sections, smears, polished cross-sections, films on a substrate
SEM images taken at 0° tilt, with the surface facing the detector
Features many pixels across, and well above the microscope's resolution
Images exported at full resolution with their original scale bar
Files whose pixel size was written by the instrument and not re-saved since
Expect errors
Tilted specimens. A surface tilted by an angle θ looks shorter by cos θ in the tilt direction: 13 % short at 30°, 29 % short at 45°. Unless your software applied tilt correction, measure along the tilt axis or correct by hand.
Features a few pixels wide. A 6 px fiber measured to ±1 px is ±17 %. Take a higher-magnification image instead of zooming the screen.
Anything near the resolution limit. A light microscope cannot separate details closer than about 0.2 µm. A 0.3 µm "width" on a brightfield image is mostly blur.
Spheres seen in a section. A slice rarely passes through the middle, so circles in a section are smaller than the particles that made them — on average about 79 % of the true diameter.
Depth. A fiber running toward or away from you looks shorter than it is, and a pore's depth cannot be read at all. A flat image gives in-plane lengths only.
A bar that no longer matches. Zoom changed on a stereo microscope after the bar was set, the wrong objective selected in the software, an image stretched in a slide deck — the bar is drawn, not measured.
Tilt shortens lengths in one direction only: 13 % at 30°, 29 % at 45°.A section through a sphere shows its true diameter only when the cut passes through the center.
The check is always the same: measure one thing whose size you know — a division of a stage micrometer, a calibration grating, a bead of certified size. If that reads right, the rest can be trusted to the same percentage.
Keep the numbers with the picture
A measurement nobody can trace back to its image is a number on a sticky note. With a free account, the image and everything drawn on it are saved together as a project.
Every shape is numbered and can carry a comment
L1, C3, S2 on the image match the same labels in the report. Add a comment to any shape — "taken 20 µm from the left edge", "second grain, etched side" — or drop a note pin anywhere on the image.
A PDF report that states where the scale came from
The report opens with a Calibration section, then lists every measurement by type, with the annotated image above them. When the scale came from the file, that section says so and names the source and the pixel size — the line a reviewer or a colleague will ask about.
Share a project by link, or invite one person by email. They see the image, the scale and every measurement, and can pan and zoom — but cannot change anything.
Create several projects from several images at once; a file that states its pixel size, or shows a scale bar, is calibrated on the way in just as a single image is. Saved presets carry your colors, rounding and number format into each new project, so a series looks the same from first frame to last.
Saving projects and exporting need a free account with a verified email, and there is a monthly allowance of exports. Spreadsheet and raw-data exports — XLSX, and CSV with unrounded values — are PRO. XLSX · CSV
Amplia or ImageJ? Use both for what they are good at
ImageJ and its Fiji distribution are the standard for scientific image analysis, they are free, and nothing here replaces them. Amplia covers the part of the job that does not need them.
Reach for Amplia when
You need a handful of lengths, diameters or areas, placed by eye
You are on a machine where you cannot install software — a shared lab PC, a tablet, a Chromebook, a phone
You want to send someone a link to the measured image instead of a screenshot
You are teaching, and the class should be measuring in five minutes, not installing
The image came from an Olympus cellSens or JEOL SEM system and should simply open at the right scale
Reach for ImageJ / Fiji or your microscope's own software when
You need thresholding, segmentation or automatic particle counting over hundreds of objects
You measure intensity — fluorescence, densitometry, line profiles. Amplia measures geometry only
You work with z-stacks, time series or multichannel images
Your data is in a proprietary format (.czi, .nd2, .lif, .dm3). Export a TIFF, PNG or JPEG first to open it here
The analysis has to be scripted and repeated identically across a dataset
Amplia converts a TIFF for display as it loads. That is fine for geometry; keep your original file for anything that depends on pixel values.
Frequently asked questions
01How do I measure something on a microscope image with a scale bar?
Open the image, choose the Calibration tool, click the two ends of the scale bar and enter its value and unit. Then draw: a Line for a length, a Circle for a diameter, a Polygon for an area. Every value is converted from pixels using the bar you marked.
02Which microscope files open already calibrated?
Images from Olympus cellSens and from JEOL scanning electron microscopes, when the file still carries the pixel size the instrument's software wrote. Amplia shows the value it applied and marks it as unverified, so check one known feature. Nikon Digital Sight cameras are recognized, but their files carry no pixel size.
03My image is from a Zeiss, Leica, Nikon or Thermo Fisher microscope. Can I still measure it?
Yes. Export it as TIFF, PNG or JPEG with the scale bar switched on, and calibrate on the bar with two clicks — or create a project from it and let Amplia read the bar. Only the automatic scale from the file is limited to certain instruments; measuring is not.
04Which units can I use?
Any metric length from picometers to kilometers — including micrometers (µm), nanometers (nm) and angstroms (Å) — and the matching areas, such as µm² and nm². With automatic unit conversion on, each value is shown in the most readable unit.
05Can I work out the real size from the magnification instead?
Only with care. Real size = size in the image ÷ magnification, but "the image" there is the eyepiece view or a print of one fixed size, not a file on a screen. Once an image is resized, the magnification figure is wrong and the scale bar is still right. Calibrate on the bar.
06Does Amplia count cells or particles automatically?
No. Every shape is placed by you. The magic wand outlines one object per click, and a grid in real units helps you count by eye, but there is no automatic detection or counting.
07How accurate are the measurements?
As accurate as the scale and the image allow. The arithmetic is exact; the limits are how precisely the scale bar was marked, how many pixels the feature covers, and whether the specimen is flat and untilted. A 1 px error on a 100 px scale bar is a 1 % error in every measurement.
08Is my unpublished image uploaded anywhere?
Not while you work as a guest: the image is opened and measured in your browser and never leaves your device. If you save it as a project in a free account, the image is stored so you can reopen and share it, and you can delete it permanently at any time.
09Can I use it on a phone or tablet at the microscope?
Yes. It runs in the browser on phones and tablets, and can be installed like an app. Press and hold to place a point with a magnified strip for precision. A photo taken through the eyepiece has no scale bar of its own, so use an eyepiece with a reticle and calibrate on its divisions.
10Is it free?
Measuring is free and needs no account, including the Line, Polyline, Polygon, Circle, Angle and Comment tools and the automatic scale from the file. A free account adds saved projects, the scale-bar reading, sharing, and PDF and image exports. A few tools and exports are marked PRO.
11Is this a replacement for ImageJ?
No. It replaces opening ImageJ for a few quick manual measurements. For automated analysis, intensity measurements, stacks or scripted workflows, ImageJ, Fiji or your microscope vendor's software remain the right tools.