esfrChart

Imatest edge spatial frequency response (eSFR) test chart

Description

An esfrChart object stores the positions and measurements of regions of interest (ROIs) of Imatest^® edge spatial frequency response (eSFR) test charts [1], [2].

The esfrChart object supports the Enhanced and Extended versions of the eSFR test chart. These test charts are based on the ISO 12233:2014 standard test chart, and they have extra visual features such as color ROIs and additional slanted edge ROIs. The esfrChart object also accepts versions of the Enhanced and Extended eSFR test charts with additional background wedges.

Creation

Syntax

chart = esfrChart(A)

chart = esfrChart(A,"Sensitivity",s)

chart = esfrChart(A,"RegistrationPoints",p)

chart = esfrChart(___,Name=Value)

Description

chart = esfrChart(A) creates an esfrChart object from an image of a test chart, A. The esfrChart object performs automatic detection of the chart position and style.

chart = esfrChart(A,"Sensitivity",s) creates an esfrChart object using sensitivity s during the automatic chart detection.

chart = esfrChart(A,"RegistrationPoints",p) creates an esfrChart object by specifying the position, p, of registration points.

chart = esfrChart(___,Name=Value) refines the automatic chart detection using one or more name-value arguments.

Input Arguments

A — Test chart image

m-by-n-by-3 numeric array | m-by-n numeric matrix

Test chart image, specified as an m-by-n-by-3 numeric array representing an RGB image or an m-by-n numeric matrix representing a grayscale image. This argument sets the Image property.

If you specify a grayscale image, then the esfrChart object simulates a color image by replicating the pixel intensity values across the three color channels. In this case, color measurements returned by the measureColor function are meaningless.

Some measurement functions expect that test chart data is in the sRGB color space. If your image is not in the sRGB color space, you can convert it to the intermediate L*a*b* color space using rgb2lab and then to the sRGB color space using lab2rgb.

Data Types: single | double | uint8 | uint16

s — Sensitivity

0.5 (default) | numeric scalar in the range [0, 1]

Sensitivity of chart detection, specified as a numeric scalar in the range [0, 1]. If you set a high sensitivity value, then the esfrChart object detects more points of interest with which to register the test chart image.

Data Types: single | double

p — Position of registration points

4-by-2 numeric matrix

Position of registration points used to orient the image, specified as a 4-by-2 numeric matrix. The four rows correspond to the top-left, top-right, bottom-right, and bottom-left registration points, respectively. The two columns represent pixel coordinates in [x, y] format. This argument sets the RegistrationPoints property.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: chart = esfrChart(A,Style="Extended") specifies that the style of the test chart is the Extended version.

Before R2021a, use commas to separate each name and value, and enclose Name in quotes.

Example: chart = esfrChart(A,"Style","Extended") specifies that the style of the test chart is the Extended version.

Style — Style of test chart

"Extended" | "Enhanced" | "WedgeExtended" | "WedgeEnhanced"

Style of test chart, specified as "Enhanced", "Extended", "WedgeEnhanced", or "WedgeExtended". If you do not specify a chart style, then by default the esfrChart object estimates the chart style based on the number and position of points of interest. This argument sets the Style property.

Data Types: char | string

CameraParameters — Camera parameters

cameraParameters object

Camera parameters used to compensate for distortion, specified as a cameraParameters (Computer Vision Toolbox) object. Use of this argument requires Computer Vision Toolbox™.

RefinePoints — Refine position of slanted edge ROIs

true or 1 (default) | false or 0

Refine the position of slanted edge ROIs, specified as a numeric or logical 1 (true) or 0 (false).

The esfrChart object first performs an initial estimate of ROI positions with respect to the registration points, RegistrationPoints. When you specify RefinePoints as true, the object then refines the slanted edge ROI positions using localized information about the image content. When false, the object does not refine the slanted edge ROI positions.

Properties

Image — Test chart image

m-by-n-by-3 numeric array

Test chart image, specified as an m-by-n-by-3 numeric array.

Data Types: single | double | uint8 | uint16

SlantedEdgeROIs — Position and intensity values of slanted edges

60-by-1 vector of structures

Position and intensity values of slanted edges, specified as a 60-by-1 vector of structures. Each element in the vector corresponds to one ROI and contains these fields:

Field	Description
ROI	Spatial extent of the ROI, returned as a 1-by-4 vector of the form [X Y Width Height]. X and Y are the coordinates of the top-left corner of the ROI. Width and Height are the width and height of the ROI, in pixels. ROI is of data type double.
ROIIntensity	Array of intensity values within the ROI, in RGB format. The array has dimensions Height-by-Width-by-3. The data type of ROIIntensity matches the data type of the Image property.

The esfrChart object excludes some slanted edge ROIs in the "Enhanced" and "WedgeEnhanced" style test charts.

• When the chart style is "Enhanced", the esfrChart object excludes four ROIs, with indices 1, 19, 41, and 59.

• When the chart style is "WedgeEnhanced", the esfrChart object excludes twelve ROIs, with indices 1, 2, 4, 18, 19, 20, 41, 42, 44, 58, 59, and 60.

For the excluded ROIs, the value of the ROI field is [NaN NaN NaN NaN] and the value of the ROIIntensity field is an empty array, [].

GrayROIs — Position and intensity values of gray patches

20-by-1 vector of structures

Position and intensity values of gray patches, specified as a 20-by-1 vector of structures. Each element in the vector corresponds to one ROI and contains these fields:

Field	Description
ROI	A 1-by-4 vector specifying the spatial extent of the ROI. The vector has the form [X Y Width Height]. X and Y are the coordinates of the top-left corner of the ROI. Width and Height are the width and height of the ROI, in pixels. ROI is of data type double.
ROIIntensity	Array of intensity values within the ROI, in RGB format. The array has dimensions Height-by-Width-by-3. The data type of ROIIntensity matches the data type of the Image property.

ColorROIs — Position and intensity values of color patches

16-by-1 vector of structures

Position and intensity values of color patches, specified as a 16-by-1 vector of structures. Each element in the vector corresponds to one ROI and contains these fields:

Field	Description
ROI	A 1-by-4 vector specifying the spatial extent of the ROI. The vector has the form [X Y Width Height]. X and Y are the coordinates of the top-left corner of the ROI. Width and Height are the width and height of the ROI, in pixels. ROI is of data type double.
ROIIntensity	Array of intensity values within the ROI, in RGB format. The array has dimensions Height-by-Width-by-3. The data type of ROIIntensity matches the data type of the Image property.

RegistrationPoints — Position of registration points

4-by-2 numeric matrix

Position of registration points used to orient the image, specified as a 4-by-2 numeric matrix. The four rows correspond to the top-left, top-right, bottom-right, and bottom-left registration points, respectively. The two columns represent pixel coordinates in [x, y] format.

Data Types: double

ReferenceGrayLab — Reference values of gray ROIs

20-by-3 numeric matrix

Reference values of gray ROIs in the CIE 1976 L*a*b* color space, specified as a 20-by-3 numeric matrix. The three columns contain the L*, a*, and b* values of the gray patches, respectively. The rows contain the reference intensities of the 20 gray ROIs, in the same sequential order.

Note

The esfrChart object includes default CIE 1976 L*a*b* values for the gray ROIs. However, the actual reference values can vary depending on several factors, such as print quality.

Data Types: double

ReferenceColorLab — Reference values of color ROIs

16-by-3 numeric matrix

Reference values of color ROIs in the CIE 1976 L*a*b* color space, specified as a 16-by-3 numeric matrix. The three columns contain the L*, a*, and b* values of the color patches, respectively. The rows contain the reference intensities of the 16 color ROIs, in the same sequential order.

Note

The esfrChart object includes default CIE 1976 L*a*b* values for the color ROIs. However, the actual reference values can vary depending on several factors, such as print quality. Accurate reference color values result in more faithful color reproduction measurements.

Data Types: double

Style — Style of test chart

"Enhanced" | "Extended" | "WedgeEnhanced" | "WedgeExtended"

Style of test chart, specified as "Enhanced", "Extended", "WedgeEnhanced", or "WedgeExtended".

Object Functions

measureSharpness	Measure spatial frequency response using test chart
measureChromaticAberration	Measure chromatic aberration at slanted edges
measureNoise	Measure noise of test chart
measureColor	Measure color reproduction using test chart
measureIlluminant	Measure scene illuminant using test chart
displayChart	Display test chart with overlaid regions of interest

Examples

Create eSFR Chart Object from Test Image

Read an image of an eSFR chart into the workspace. Display the image.

I = imread("eSFRTestImage.jpg");
imshow(I)
title("Captured Image of eSFR Chart")
text(size(I,2),size(I,1)+15,["Chart courtesy of Imatest",char(174)], ...
    FontSize=10,HorizontalAlignment="right");

Create an esfrChart object using the chart image.

chart = esfrChart(I)

chart = 
  esfrChart with properties:

                 Image: [1836×3084×3 uint8]
       SlantedEdgeROIs: [60×1 struct]
              GrayROIs: [20×1 struct]
             ColorROIs: [16×1 struct]
    RegistrationPoints: [4×2 double]
                 Style: 'Extended'
      ReferenceGrayLab: [20×3 double]
     ReferenceColorLab: [16×3 double]

Display the imported eSFR chart. Regions of interest (ROI) are highlighted and labeled.

displayChart(chart)

The chart is imported correctly. All 60 slanted edge ROIs (labeled with green numbers) are visible and centered on appropriate edges. 20 gray patch ROIs (labeled in red) and 16 color patch ROIs (labeled in white) are visible and are contained within the boundary of each patch.

Create eSFR Chart Object Using Specified Registration Points

Create an esfrChart object by specifying the coordinates of the four registration points. Registration points are located at the center of the black-and-white checkered circles.

Read an image of an eSFR chart into the workspace.

I = imread("eSFRTestImage.jpg");

Display the image and configure it to collect four registration points.

figure
imshow(I)
[X, Y] = ginput(4);

Click the registration points in this order: top-left, top-right, bottom-right, bottom-left.

Create an esfrChart object, specifying the four registration points. Display the imported eSFR chart. Regions of interest are highlighted and labeled. The registration points appear in red.

chart = esfrChart(I,RegistrationPoints=[X, Y]);
displayChart(chart);

Tips

• For accurate and reliable results, acquire an image of the test chart according to standard specifications outlined in the ISO standard and by the manufacturer [2], [3]. As a simple guideline, align the chart horizontally on a light background. Cover over 90% of the field of view with the chart, but ensure that the top and bottom edges of the chart are still visible. For reliable measurements, set the minimum image width to at least 500 pixels.

• You can capture an image of the Extended eSFR test chart at the full 16:9 aspect ratio, or at an aspect ratio of 3:2 or 4:3, as specified on the chart.

• To ensure that the chart is properly imported, visually verify the test chart image using the displayChart function.

• The size and position of the estimated ROIs can vary slightly depending on your operating system and machine hardware.

References

[1] Imatest. "Esfr". https://www.imatest.com/mathworks/esfr/.

[2] Using eSFR ISO Part 1. URL: https://www.imatest.com/docs/esfriso_instructions.

[3] ISO 12233:2014. "Photography – Electronic still picture imaging – Resolution and spatial frequency responses." International Organization for Standardization; ISO/TC 42 Photography. URL: https://www.iso.org/standard/59419.html.

Version History

Introduced in R2017b

R2021a: esfrChart object now refines slanted edge ROI positions

Starting in R2021a, by default the esfrChart object refines the position of slanted edge ROIs based on localized image content. In previous releases, the esfrChart object did not refine the position of the slanted edge ROIs after the initial estimate of the position based on the position of the registration points. The refinement results in more precise ROI positions.

To replicate the previous behavior, specify the RefinePoints name-value argument as false when creating an esfrChart object.

R2021a: Read and analyze additional versions of Imatest eSFR test charts

The esfrChart object can detect and analyze the Enhanced, Wedge Enhanced, and Wedge Extended versions of the Imatest eSFR test chart.

You can specify camera parameters to compensate for distortion using the CameraParameters name-value argument.