How to Measure Camera Sharpness with a Siemens Star
There are various methods for measuring the optical resolution or sharpness of a camera. One of the simplest methods is sharpness measurement using a Siemens star. This pattern consists of bright “spokes” on a dark background that radiate from a common center, becoming wider as they extend outward. In a perfect optical system, the spokes meet sharply at the center. In an imperfect system, the spokes blur at a certain distance from the center.
To manually determine sharpness, you can measure the radius at which the Moiré pattern begins to appear clearly and apply it to the following formula:
Spatial frequency = number of lines / (2 x Radius x pi)
However, since hyperspectral light field cameras consist of dozens or even hundreds of lenses, an automated and standardized solution is required.
How to Measure Camera Sharpness with the Slanted Edge Method
Defined in ISO 12233, the slanted edge method measures the Modulation Transfer Function (MTF) as a function of the image’s spatial frequency. In this method, a knife-edge target is captured by the camera, and a user-defined region of interest is selected. The process is detailed here:
We have adapted this method for use with our hyperspectral cameras. This method is employed in our internal quality control processes to ensure the sharpness of all our cameras. For each lens of the light field hyperspectral camera, we first calculate the Edge Spread Function (ESF) and then derive the MTF from it. From the MTF, we can determine the camera’s maximum theoretical resolution in micrometers.
Results
The Slanted Edge Sharpness measurement is a standard tool for quality control in all Cubert cameras. This method ensures maximum performance for all our devices. Attached are typical results from our Ultris X20 model. We achieve very high imaging quality with a resolution of 7.2 µm, which is comparable to the pixel pitch of 6.4 µm for the ULTRIS X20 sensor.
Figs 3 and 4: For each channel, we calculate the Edge SpreadFunction, which generates the Modulation Transfer Function in lines per mm.
Fig 5: Plot of the measured maximum objective resolution of a sample ULTRIS X20. The pixel size of the camera is 6.4 micrometers, and with a mean resolution of 7.26 micrometers, the objective delivers near-perfect image reproduction
About the Author
Dr. René Heine is the Co-Founder and CEO of Cubert GmbH, a leader in real-time spectral imaging. Since founding the company in 2012, René has been instrumental in shaping Cubert’s technological direction and growth. He holds a Doctor of Physics degree from the University of Ulm, where he graduated magna cum laude, and completed his diploma thesis at Harvard Medical School. René’s deep expertise in physics and his vision for cutting-edge imaging technologies drive Cubert’s innovations and advancements in hyperspectral solutions.
