Modulation Transfer Function (MTF) Measurements

Owned by Former user (Deleted)
Jun 19, 2013
3 min read

In IQWare, Modulation transfer function (MTF) is measured using slanted edge chart. MTF measurement interface can be accessed through MTF tab. The capture buttons (DevWare Image and Load Image) are on the top. The loaded chart image is displayed on the left (Chart Image group), and the results are displayed on the right (MTF Plot group). Text results are displayed on the bottom.

MTF measurement uses Slanted Edge chart (ISO 12233 chart). Any slanted edge (positive slope) with sufficient contrast can also be used. An image with Slanted Edge chart is needed for measurement or it can be captured in DevWare through IQWare interface.

In the example below, ISO12232 chart with possible slanted edge regions selected are shown.

Basic MTF Measurement Steps

Basic measurement can be performed following these steps:

  • Load the image to be measured using one of the two ways below.

    • Capture from DevWare by clicking the DevWare Image button, DevWare is loaded and the following Setup Sensor dialog will show. DevWare settings might need adjustments to get proper streaming data. Compose the scene with a Macbeth Chart at the front of camera. Once the scene is composed in DevWare, press the OK button in the Setup Sensor dialog. The dialog will disappear and the DevWare image will be captured and loaded in IQWare. (Please note that once DevWare is running, the Setup Sensor dialog will not be prompted again. Image capturing is done by clicking the DevWare Image button again.)

    • Load existing image file by clicking the Load Image button. A file dialog will be promopted to browse and select the image file for measurement.

  • Set up the regions on interest using "Make Region Selection" button . Regions need have contrast of at least 20% and slanted edge should have slope between 3-7 degrees. Region can be resized or moved after creation. Use button to activate the mode.

    Existing regions can be loaded using  .(Note that loading regions from region file will replace existing regions. It is NOT append operation). The regions can be saved for future use by using .

  • Select the value or input an known gamma value for this image. (Mismatched Gamma value will generate invalid results). The calculation is done on linear image and so any non-linear (Gamma not equal to unity) will be linearized.)

  • Click the Calculate MTF button to start MTF measurement.This may take a few seconds depending on the number of regions to be processed. A progress bar will be shown to users during the processing. The MTF chart for the current active region will be plotted on the right MTF Plot window. Each of RGB channels can be toggled on/off by clicking the legend on the right. The active region can be selected by clicking on the region.

  • Report will be displayed on the bottom. If "Report only selected region" is checked, only results from active region is reported. Otherwise, the results from all regions are reported. If "Auto re-run calculation" is checked, "Calculate MTF" will become "checkable button". In this mode, calculation will be refreshed for the new regions or resized or moved regions as long as "Calculate MTF" is depressed/ activated. In the report, distance from center, overshoot, MTF50 and MTF10 frequencies are also reported. Overshoot is calculated as maximum MTF response above 1.0. Distance from center is the distance in pixels between mid-point of image and mid-point of the region of interest. Slope of the the edge is reported for each region. The regions with the slopes beyond defined range (2-7 degrees) are have red background for their slope values.

  • Click the Save to HTML report (Shortcut key: Ctrl + S) menu under Save button on the top left corner to save the measurements and MTF chart to select destination. The report saved will include results of all regions of interest, extracted image of the region and plots for each region. If any region faield to generate valid results, error information is provided with region information(including the extracted region image).