U.S. flag An official website of the United States government

U.S. Food and Drug Administration

Contrast Transfer Function Analysis Tool for Head Mounted Displays

Catalog of Regulatory Science Tools to Help Assess New Medical Devices 

This regulatory science tool (RST) is a laboratory method that calculates the Contrast Transfer Function (CTF) of Augmented Reality (AR) and Virtual Reality (VR) head mounted displays (HMDs) by quantifying the Michelson contrast as a function of spatial frequency for horizontal or vertical grille patterns displayed on the HMD.

Technical Description

Virtual image contrast and spatial resolution are critical aspects of image quality for many medical applications of AR and VR HMDs. These image quality parameters can be characterized using the Michelson contrast of grille patterns, which are described in IEC632145-20-20 and ICDM IDMS standards documents. The measurements are repeated for grille patterns with different periodicities to characterize the image contrast as a function of spatial frequency or feature size.. The Michelson contrast for the grille patterns can be plotted as a function of the spatial frequency of the grille pattern to show the dependence of the image contrast on the spatial frequency of the virtual image content. The resulting plot is known as the contrast transfer function (CTF) and is commonly used to visualize the performance of AR and VR HMDs.

The CTF analysis tool computes the Contrast Transfer Function (CTF) for AR and VR head mounted displays (HMDs) by quantifying the Michelson contrast as a function of spatial frequency for horizontal or vertical grille patterns displayed on the HMD. We developed a general method that can be applied to a wide variety of HMDs, encompassing different design and display technologies to generate reproducible CTFs. The analysis tool we developed automates the CTF analysis and calculations with minimal user interaction (Fig. 1).

CTF analysis tool

Figure 1. This figure shows the graphical user interface of the CTF analysis tool, displaying the result curve generated from the example included in the tool package.

To compute the CTF of the HMD, a grille pattern with equal-width white and black bars, oriented either vertically or horizontally, is displayed on the HMD screen. The varying widths of the bars correspond to different spatial frequency values in cycles per degree. This allows for Michelson contrast measurements at these spatial frequencies. The acquired images of the displayed grille patterns are then used as input for the CTF analysis tool.

As part of the tool’s functionality, the user may need to smooth the data to resolve issues related to the subpixel structure caused by the display technology used in the HMD. In such cases, the user should acquire a full screen flat white images to compute the subpixel spacing and determine the optimal filter size for smoothing the data.

The user must also provide the accurate horizontal field of view (FOV) of the camera used to acquire the images. This value is entered as input into the tool. Additionally, the user must select the direction of the grille pattern displayed on the HMD, either "Vertical" or "Horizontal.

The output results of the CTF analysis tool are listed as follows:

  • The CTF curve shows the contrast values as a function of spatial frequency (cycles per degree).
  • Standard deviation of luminance/signal from bright (white pixels) and dark (black pixels) areas displayed in the HMD.
  • Average luminance/signal from bright (white pixels) and dark (black pixels) areas displayed in the HMD.
  • CTF cutoff point at the highest spatial frequency value allowed by resolution and/or rendering limits of the HMD.
  • Data table of all outputs saved in CSV format. 

Intended Purpose 

The CTF analysis tool was developed as a general method to generate reproducible CTFs  that can assist in assessing the performance of a wide variety of HMDs, encompassing different design and display technologies. The analysis tool automates the CTF analysis and calculations with minimal user interaction. 

Testing

The recommended bench setup, including the light measurement device (LMD), motion controls, and alignment procedures, are described in detail in ICDM IDMS standards document. 

The CTF analysis tool requires images with a resolution greater than 150 pixels per degree (ppd). If a camera with a lower resolution is used, an error message will appear. The camera used to test the CTF tool has a resolution of 265 ppd. The camera settings are the camera's gain set to 0 and the gamma value is set to 1. Additionally, the camera’s integration time must be set to maximum, but care should be taken to avoid pixel saturation to average over several HMD frames. For shorter integration times, the camera integration should be a multiple of the inverse of the HMD frame rate. 

The CTF analysis tool was tested across various parameters, including pattern direction, eye rotation angles, display and optical design, cameras, and the rendering techniques. This aims to develop an automated tool to assist developers evaluating CTFs. The method has been extensively tested on multiple HMDs include:

  • HTC Vive Pro
  • HTC Vive XR Elite
  • Meta Quest 2
  • Meta Quest 3
  • Magic Leap 2

The resulting CTF plots were compared with previous research findings to ensure that the developed tool’s processing steps align precisely with the standard image processing procedures. We repeatedly computed the CTF for different parameters, including various filter sizes and integration times. The tests involved repeating measurements at different locations within the FOV of the HMD, across different eyepieces of each HMD, and on a conventional monitor placed at the virtual focal plane.

Limitations

Limitations of the CTF Analysis Tool:

  • The tool is only compatible with AR and VR HMDs.
  • It supports only cameras with a resolution greater than 150 pixels per degree to ensure the camera has a higher angular resolution than the HMD and sufficient spatial sampling for the test patterns on the HMD.
  • The tool can only process acquired images that feature vertical or horizontal grille patterns.
  • The accuracy of the results depends on the field of view (FOV) value; therefore, the FOV value must be accurate and error within ±2% is acceptable.
  • The minimum acceptable FOV value is 2 degrees.

Supporting Documentation

  1. Contrast Transfer Function Analysis Tool GitHub
  2. IEC 63145-20-20. Eyewear Display—Part 20-20: Fundamental Measurement Methods Image Quality. Geneva, Switzerland: IEC; 2019.
  3. International Committee for Display Metrology (ICDM). Information Display Measurements Standards SID; 2022.
  4. FDA Recognized Consensus Standards: Medical Devices, https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfStandards/search.cfm
  5. RST Toolkit for Evaluation of Head Mounted Display Image Quality, https://cdrh-rst.fda.gov/toolkit-evaluation-head-mounted-display-image-quality

Contact

Tool Reference