Imatest Version 24.1 Beta

Imatest 24.1 Beta is available below

General Improvements

• Registration Mark Detection: Now includes an alternative machine learning/computer vision (ML/CV) registration mark detection method to aid with automatic target identification and ROI placement. This new detection method is available for the following analyses: Random/Dead Leaves, eSFR ISO 12233:2017, eSFR ISO 12233:2023, Star Chart, 36-Patch Dynamic Range, Contrast Resolution, ISO 14524, and ISO 15739.
• Improved Handling of DNG files: Added ability to read non-standard bit depth DNG files using direct read option, and Incorporated Adobe DNG SDK into Libraw DNG read option for improved processing according to DNG specification. There is a new dropdown option for the DNG Read method selection in the Options II window, to allow the choice between the original Libraw processing and the new Adobe SDK processing.
• Checkerboard: Now includes a High-distortion option for checkerboard detection, providing improved detection robustness for highly distorted images.
• Rescharts and Color/Tone: Now includes the ability to hide the text overlaid on select plots as an option under the Display dropdown, located at the top of Rescharts and Color/Tone Interactive windows.
• Color/Tone and Stepchart: The functionality of the saturation level setting has been modified to allow for better handling of images with nonstandard bit depth or saturation level offsets.

Image Information Metrics — from slanted edges

The image information metrics have expanded from the metrics that were available in releases 23.1 and 23.2. The earlier releases included edge-based information capacity and noise image-based information capacity, the latter which uses the calculated noise power spectrum (NPS) as part of the information capacity equation. The NPS also enables additional metrics which are being launched in 24.1 Beta.

• SNRi and Edge SNRi  (Independent Observer Signal-to-Noise Ratios). Measures of the detectability of objects and edges.
• Matched filter design  Procedure for designing electronic filters that optimize object and edge detection metrics, SNRi and Edge SNRi.
• Slanted edge SFR interpolated binning method for reducing spurious high-frequency signals from SFR curves. Reduces sawtooth artifacts in motion blur analysis and improves the reliability of MTF10 measurements, but is not recommended for general use because it lowpass-filters the SFR.

Refer to this documentation, for instructions on using the new options for information capacity calculations, and more information on the new measurements.

Stray Light (Flare)

• Now includes support for ITU-R BT.709-6, ITU-R BT.2020-2, and MATLAB luminance calculations.
• Now includes the number of saturated pixels per analysis frame.
• Now includes a maximum measurable stray light calculation.

Main Window

• Now includes the ability to annotate images with Custom Metadata.
• Now includes improved General Settings search with highlighting.

Known Issues

This is a Beta version of the software, major issues are listed below:

• Crash occurs when using Registration Mark Detection on a Star Chart
• eSFRiso charts ROI not auto detected if “machine vision autodetect” selected on SFRreg settings