HR-TEM Image Analysis Platform

ChemFitX provides an online HR-TEM image analysis workflow for researchers working with high-resolution transmission electron microscopy images of crystalline and nanostructured materials. The platform combines calibrated lattice-fringe d-spacing measurement, multi-fringe averaging, lattice intensity-profile analysis, reciprocal-space FFT inspection, measurement statistics and export tools in one browser-based workspace.

Upload an HR-TEM image, calibrate the image from a visible scale bar or a known pixel calibration, and then measure lattice fringes directly on the micrograph. Advanced tools can estimate periodicity from independent line profiles using FFT and autocorrelation, inspect a local 2D FFT in reciprocal space, compare measured d-spacing with a user-supplied reference value, and summarize repeated measurements statistically.

The microscopy image is read and analyzed locally in your web browser by this application. The uploaded image is not sent to ChemFitX servers during the analysis.

What the HR-TEM Image Analysis Platform Can Do

Import HR-TEM images in TIFF/TIF, PNG, JPG/JPEG and BMP formats.
Calibrate image scale by measuring a visible scale bar and entering its known length.
Use direct unit-per-pixel calibration when a validated pixel size is already known.
Measure lattice-fringe d-spacing using either a single fringe interval or a multi-fringe average.
Assign an optional hkl or other text label to individual d-spacing measurements.
Report line length, number of fringe intervals, d-spacing, reciprocal spacing and line orientation.
Estimate measurement uncertainty from user-defined endpoint and scale-calibration uncertainty inputs.
Calculate a relative strain percentage when a reference d-spacing is supplied.
Draw independent lattice-profile lines and estimate periodicity using one-dimensional FFT and autocorrelation.
Analyze several profile lines separately and compare their periodicity estimates.
Compute a local two-dimensional FFT from a selectable square image region.
Inspect reciprocal-space position, reciprocal magnitude g, real-space periodicity d and angle for selected FFT spots.
Summarize repeated d-spacing measurements using count, mean, median, standard deviation, minimum and maximum values.
Display a d-spacing distribution histogram.
Adjust brightness, contrast and threshold for visualization without replacing the original pixels used by the quantitative profile and FFT routines.
Pan, zoom, rotate, flip, center and inspect the image interactively.
Customize measurement-line color, transparency, width, line style and arrow style.
Export an annotated PNG image and export numerical results to Excel, CSV or JSON.

Supported HR-TEM Images and Analysis Requirements

Supported image formats

The platform accepts TIFF/TIF, PNG, JPG/JPEG and BMP images. For quantitative lattice measurements, use the highest-quality image available and avoid unnecessary resaving, resizing or lossy compression before analysis whenever the original microscopy image is available.

A usable spatial calibration

Physical d-spacing values require a valid spatial calibration. ChemFitX can obtain that calibration from a visible scale bar or from a known unit-per-pixel value. If the image is not calibrated, measurements remain in pixels rather than physical length units.

Visible lattice fringes

The selected image region should contain resolvable lattice fringes with sufficient contrast and a meaningful number of repeated periods. Measurements become less reliable when fringes are strongly curved, blurred, overlapping, heavily distorted or close to the image-resolution limit.

Scale Calibration for HR-TEM Measurements

Scale-bar calibration

To calibrate from a printed scale bar, select the scale-calibration tool and drag a line exactly from one end of the visible scale bar to the other. Enter the numerical value printed on the scale bar and choose its unit. ChemFitX converts the measured image-pixel distance into a physical unit-per-pixel calibration.

The calibration line is measured in the original image coordinate system, so changing the zoom level does not alter the calibration. If the scale-bar value or unit was entered incorrectly, it can be corrected and the scale reset without redrawing the reference line.

Direct pixel calibration

If a validated pixel size is already known from the microscopy acquisition or image metadata, use the direct pixel-calibration option and enter the physical length represented by one pixel. Supported calibration units include nanometers, angstroms, picometers and micrometers per pixel.

Why calibration accuracy matters

Every calibrated real-space and reciprocal-space result depends on the spatial calibration. A scale-bar line that does not match the full printed bar, an incorrect scale value, or an image that has been resized after the scale bar was created can propagate directly into d-spacing and reciprocal-space values.

Lattice d-Spacing Measurement

Single-fringe measurement

In single-fringe mode, draw a line across one lattice-fringe interval. The measured line length is converted from pixels to the selected physical unit using the current calibration and is reported as the d-spacing.

Multi-fringe averaging

For a more stable manual estimate, multi-fringe mode lets you draw across several repeated lattice intervals and enter the number of intervals crossed. ChemFitX calculates d as the total calibrated line length divided by the number of fringe intervals. Averaging several periods can reduce the influence of small endpoint-placement errors when the fringe family is sufficiently uniform.

Orientation and reciprocal spacing

Each d-spacing line also has an image-plane orientation angle. The platform additionally reports the reciprocal value 1/d in the reciprocal unit corresponding to the measured real-space unit. These values describe the selected image measurement and should be interpreted together with the image geometry and calibration.

Optional hkl labels

An hkl label can be entered before a measurement so that a user-assigned plane label is stored with the result and included in exports. ChemFitX does not determine or validate the crystallographic hkl assignment automatically; the label should come from an independent crystallographic interpretation.

Estimated Measurement Uncertainty

The optional uncertainty controls provide an estimated uncertainty for each manual d-spacing value from two user-defined components: endpoint uncertainty in pixels and scale-calibration uncertainty in percent. The calculation propagates these specified contributions into the reported d-spacing.

This estimate is intended as a practical measurement aid rather than a complete electron-microscopy uncertainty budget. It does not automatically include specimen drift, local lattice curvature, microscope distortion, aberrations, sampling limitations, image-processing history, specimen tilt or uncertainty in the crystallographic reference itself.

Reference d-Spacing and Strain-Assisted Interpretation

A reference d-spacing can be entered in nm, Å, pm or μm. When a valid reference value is present, ChemFitX reports the relative difference as strain percentage using ε = (d − dref) / dref × 100%.

This result is a local d-spacing comparison relative to the reference supplied by the user. It should not be treated as a complete strain-tensor measurement or an automatic geometric-phase-analysis result. Reliable strain interpretation requires an appropriate reference, careful control of imaging geometry and consideration of specimen orientation, distortion and other experimental factors.

Lattice Fringe Intensity Profile Analysis

The profile-analysis module uses independent guide lines that are separate from the manually drawn d-spacing measurements. Draw a profile line approximately perpendicular to the lattice-fringe family of interest so the intensity signal crosses repeated bright-dark fringe periods.

Band averaging

A configurable averaging-band width samples grayscale intensity across a narrow band around the profile line. Band averaging can reduce pixel-scale noise, but a band that is too wide may mix adjacent regions or different fringe orientations.

Profile smoothing

The profile signal can be smoothed with a user-defined moving window before periodicity analysis. Smoothing should be kept conservative because excessive smoothing can suppress genuine high-frequency information or merge nearby periodic components.

FFT periodicity estimate

ChemFitX applies a one-dimensional FFT to the sampled profile and identifies a dominant non-zero spatial-frequency component within the implemented search range. The corresponding period is reported as a profile FFT d value.

Autocorrelation periodicity estimate

The same profile is also examined with autocorrelation. A dominant correlation lag is converted into a second periodicity estimate. The platform reports the percentage difference between the FFT and autocorrelation estimates so that the two independent numerical estimates can be compared.

Agreement between the two estimates can support confidence that the selected line contains a repeatable periodic signal, but agreement alone does not establish a crystallographic phase or hkl assignment.

2D FFT and Reciprocal-Space Inspection

The advanced FFT module computes a local two-dimensional fast Fourier transform from a square region of the original image. Available ROI sizes are 128 × 128, 256 × 256 and 512 × 512 pixels, subject to the dimensions of the uploaded image.

When a profile line is selected, the FFT region is centered near that profile-line midpoint. If no suitable profile line is selected, the platform can use the selected d-spacing measurement and then the image center as a fallback. An optional Hann window can be applied before the transform to reduce edge leakage.

Reciprocal-space spot inspection

After the FFT is generated, move the cursor or select spots away from the transform center to inspect reciprocal magnitude g, corresponding real-space periodicity d and reciprocal-space angle. With a valid real-space calibration, g is reported in inverse length units and d is calculated as 1/g.

FFT spot picking is an inspection and measurement aid. A selected spot does not by itself prove a crystal phase, zone axis or hkl assignment. Phase identification should use appropriate crystallographic reference data and, where needed, complementary diffraction or simulation.

d-Spacing Statistics and Distribution

Repeated manual d-spacing measurements are summarized automatically. The statistics panel reports the number of measurements, mean d-spacing, median d-spacing, standard deviation, minimum and maximum values, and mean strain when valid strain values are available.

A histogram provides a visual overview of the measured d-spacing distribution. Statistical summaries are most meaningful when the measurements represent a clearly defined sampling strategy. Combining unrelated phases, different fringe families or strongly different regions into one measurement set can produce a distribution that is difficult to interpret physically.

Image Display, Navigation and Annotation

The analysis workspace includes pan, zoom, reset, rotate, horizontal flip, vertical flip, centering and fullscreen controls. Cursor readouts provide image coordinates and grayscale information while inspecting the micrograph.

Brightness, contrast and threshold controls are intended to improve visualization. In the supplied platform, the quantitative line-profile and FFT calculations use the original uploaded image pixels rather than the display-adjusted appearance. This helps separate visual enhancement from the numerical signal used for those advanced analyses.

Measurement and calibration lines can be customized with different colors, transparency levels, widths, solid/dashed/dotted styles and optional arrow styles. These controls are useful when preparing an annotated figure or distinguishing multiple measurements.

How to Use the ChemFitX HR-TEM Image Analysis Tool

Upload an HR-TEM image in TIFF/TIF, PNG, JPG/JPEG or BMP format.
Inspect the image and confirm that the scale bar and lattice fringes are clearly visible.
Calibrate the image by measuring the full scale bar and entering its known value, or enter a validated direct pixel calibration.
Select single-fringe or multi-fringe d-spacing measurement. For multi-fringe mode, enter the number of fringe intervals crossed by the line.
If needed, enter an hkl label, reference d-spacing, endpoint uncertainty and scale uncertainty before drawing measurements.
Draw the d-spacing line approximately perpendicular to the selected lattice-fringe planes and inspect the calculated value.
Repeat measurements in representative regions when statistical evaluation is required.
Use independent profile lines for advanced periodicity analysis and compare FFT and autocorrelation estimates.
Compute a local 2D FFT when reciprocal-space inspection is useful, and select FFT spots only after confirming the real-space calibration.
Review the d-spacing statistics, histogram, measurement list, profile table and FFT spot table together with the original image.
Customize measurement lines if preparing an annotated output image.
Export the annotated PNG and/or numerical results to Excel, CSV or JSON.

Interpreting the Main HR-TEM Results

d-spacing

The reported d-spacing is the calibrated real-space periodic distance represented by the selected fringe measurement. Its physical interpretation depends on correct scale calibration, fringe selection and crystallographic context.

Multi-fringe d-spacing

For a line crossing N repeated intervals, the platform divides the total calibrated line length by N. The interval count should represent the actual number of repeated lattice periods crossed by the line; an incorrect count directly changes the calculated d value.

Orientation

The reported orientation is the angle of the measurement line within the image coordinate system. Because the d-spacing line is normally drawn across the fringe periodicity, it should be interpreted as a measurement-line orientation rather than an automatically determined crystallographic plane-normal in a fully calibrated specimen coordinate system.

Profile FFT and autocorrelation d

The profile FFT d and autocorrelation d are periodicity estimates derived from grayscale intensity variations along an independent profile line. Similar values from the two methods can indicate a consistent repeating signal, while a large difference can indicate noise, line misorientation, multiple frequencies or an insufficient profile length.

2D FFT g and d

In the local 2D FFT, the distance from the transform center corresponds to spatial frequency. After calibration, the selected spot gives reciprocal magnitude g and the associated period d = 1/g. The reported angle describes the reciprocal-space spot direction in the displayed FFT coordinate system.

Why Multi-Fringe Averaging Can Be Useful

A one-period manual measurement can be sensitive to exactly where the two endpoints are placed. When several nearly parallel and evenly spaced fringes are visible, measuring across multiple periods and dividing by the number of intervals can reduce the relative influence of endpoint placement. This is why the platform provides multi-fringe averaging as the default measurement strategy.

Multi-fringe averaging should not be used blindly across regions where the spacing changes continuously, fringes bend strongly, separate domains overlap or multiple lattice periodicities are present. In those situations, shorter local measurements or region-specific analysis may be more meaningful.

Exporting HR-TEM Analysis Results

Annotated image

The image-export function creates a PNG containing the current image orientation and the selected measurement annotations. Measurement labels can be included or omitted, and the export font family, font size and font weight can be adjusted.

Excel

The Excel export can include d-spacing measurements, calibration information, line-profile summary results and selected FFT spots when those analyses are available.

CSV

CSV export contains the manual d-spacing measurement table, including measurement ID, optional label, line length in pixels, fringe-interval count, d-spacing, unit, estimated uncertainty, reciprocal value, orientation, reference d and strain when available.

JSON

JSON export stores structured calibration information together with d-spacing measurements, analyzed line profiles and selected FFT-spot results. This format is useful when results will be parsed or reused programmatically.

Data Privacy and Browser-Based Processing

For this HR-TEM platform, the uploaded microscopy image is read and processed locally in the browser. The analysis code does not upload the image to ChemFitX servers.

Users should still follow their laboratory or institution’s own rules for unpublished images, confidential research data, shared computers, browser storage and exported files.

Scientific Scope and Limitations

The platform is an image-analysis and measurement aid; it is not a substitute for microscope calibration, crystallographic indexing or expert TEM interpretation.
Accurate physical measurements require a valid spatial calibration. A wrong scale-bar measurement or pixel calibration produces correspondingly wrong d-spacing and reciprocal-space values.
Manual d-spacing results depend on endpoint placement and, in multi-fringe mode, the correct number of lattice intervals.
The estimated uncertainty includes only the user-entered endpoint and scale-calibration contributions implemented by the platform; it is not a complete microscopy uncertainty model.
The reference-based strain value is a relative d-spacing comparison, not a full strain tensor or automatic geometric-phase-analysis map.
User-entered hkl labels are not verified automatically by ChemFitX.
FFT spot selection provides reciprocal periodicity and angle information but does not automatically identify a phase, zone axis or crystallographic plane.
Line-profile FFT and autocorrelation estimates can be affected by noise, multiple fringe families, line orientation, insufficient profile length and non-uniform lattice spacing.
Brightness, contrast and threshold adjustments change visualization, while the implemented quantitative profile and FFT routines use the original image pixels.
Specimen tilt, thickness, drift, aberrations, scan or camera distortion, image filtering, compression and acquisition conditions can influence apparent lattice-fringe measurements.
Measurements made from JPEG images can be influenced by lossy compression; use the original microscopy data when available.
Statistical summaries should combine measurements only when the sampling strategy and physical population being summarized are scientifically appropriate.

Frequently Asked Questions

What is HR-TEM image analysis?

HR-TEM image analysis examines high-resolution transmission electron microscopy images to quantify or inspect features such as lattice-fringe spacing, periodicity, orientation and reciprocal-space information. Interpretation should remain linked to microscope calibration, imaging conditions and crystallography.

Can I measure lattice d-spacing online with ChemFitX?

Yes. After calibrating the image, draw a d-spacing line across the lattice fringes. The platform can use either a single interval or the average of several fringe intervals.

Which image formats are supported?

The platform accepts TIFF/TIF, PNG, JPG/JPEG and BMP files.

Do I need a scale bar?

A visible scale bar is one way to calibrate the image. If a validated physical pixel size is already known, you can instead use direct pixel calibration.

Why should I measure several lattice fringes instead of only one?

When the same fringe family is uniform, measuring across several periods can reduce the relative effect of small endpoint-placement errors. You must enter the correct number of intervals crossed.

Does ChemFitX automatically identify hkl planes?

No. You can attach an hkl label to a measurement, but the label is user supplied. The platform does not automatically determine or validate crystallographic indexing.

What does the strain value mean?

If you provide a reference d-spacing, the platform reports the relative difference between measured and reference spacing as a percentage. It is a reference-based d-spacing comparison and not a complete strain-tensor analysis.

What is the purpose of the lattice intensity profile?

A profile line samples grayscale intensity across repeated fringes. ChemFitX estimates the dominant periodicity using FFT and autocorrelation so that the local repeating signal can be inspected independently of manual d-spacing lines.

What does the 2D FFT show?

The local 2D FFT represents spatial-frequency content in reciprocal space. After calibration, selected spots can be expressed as reciprocal magnitude g, corresponding real-space period d and angle.

Can an FFT spot identify a crystal phase automatically?

No. Reciprocal-space spot position alone is not sufficient for automatic phase or hkl identification. Use appropriate crystallographic references, diffraction information and expert interpretation.

Do brightness and contrast changes alter the quantitative FFT results?

In this platform, display adjustments are for visualization, while the quantitative line-profile and FFT calculations use the original uploaded image pixels.

Are my HR-TEM images uploaded to ChemFitX?

For this application, the microscopy image is read and analyzed locally in the browser and is not uploaded to ChemFitX servers by the analysis code.

Can I export my measurements?

Yes. You can export an annotated PNG image and numerical results in Excel, CSV or JSON formats. Excel and JSON can also include advanced profile and FFT results when available.

Scientific References

Williams, D. B.; Carter, C. B. Transmission Electron Microscopy: A Textbook for Materials Science. Springer.
Reimer, L.; Kohl, H. Transmission Electron Microscopy: Physics of Image Formation. Springer.
Kirkland, E. J. Advanced Computing in Electron Microscopy. Springer.