The Resolution of Lenses and Images.
What is the Definition of Resolution? And Why Does it Matter?
Resolution defines the sharpness of an image and can indicate advanced camera functions.
There are four types of resolution:
1.) Pixel Count Resolution
2.) System Resolution
3.) Lens Resolution
4.) Angular Resolution
All types of resolution are related and jointly impact computer vision.
Why Can Resolution Be So Confusing?
You're not alone! Even the experts interchangeably use the term "Resolution" when it actually refers to four different technical metrics. We ran across this misunderstanding between Optical Engineers, Image Quality Engineers, Sensor Engineering, and Computer Vision Engineers during a 2018 conversation within the IEEE P2020 working group.
The resolution of a sensor is also closely related to the CMOS sensor format type (e.g. 1/3", 1/2", etc..) which we discuss in a separate article.
Megapixels as Resolution
Pixel Count Resolution
The pixel count resolution has been popularized by the mobile phone industry, but is frequently misleading for 50MP+ cameras as the final output/viewed image rarely has 50MP+ resolution.
Definition:
- The total number of light capturing photodiode wells in a sensor.
- The total number of light emitting elements for displays and projectors.
- The total number of pixels in a digital image.
Accurate Usage:
The camera has 64 MegaPixels and creates a 12 to 16 Megapixel image.
Inaccurate Usage:
The camera is 64MP.
SFR as Resolution
What is System Resolution and Why Does it Matter?
The system resolution can characterize the quality any digital image. It is frequently used in digital camera production lines to ensure the lens, lens-to-sensor alignment, and image processing pipeline meet a minimum quality threshold. It can also be used to benchmark different cameras and image processing/compression methods. This metric includes inputs from the Lens, the Lens-to-sensor alignment, the Sensor, the ADC, and the Analog+Digital Image Processing Pipeline.
Definition:
- The Spatial Frequency Response (SFR) measures the change in contrast across a gradient in the image.
How to Calculate Systen Resolution:
Follow ISO12233 or read Imatest's website.
How to Express Systen Resolution:
Include the line-widths per picture height, the field position, and the illumination.
Accurate Usage:
The SFR is 2000 LW/PH at 25° field angle under 1000lux 5000K illumination.
Inaccurate Usage:
The MTF of the image is 2000 LW/PH.
What is The Impact of SFR on Computer Vision?
From: Vasiljevic, et. Al. “Examining the Impact of Blur on Recognition by Convolutional Networks”
MTF as Resolution
"Optical Resolution" is used by Imaging Optics Engineers and is an Input to System Resolution
The optical resolution characterizes the 'sharpness' of an optical system. The MTF applies to both focal and afocal optical systems, including microscopes, imaging lenses, binoculars, and projection optics.
Definition:
- The ability of an optical system to resolve two points.
How to Express Optical Resolution:
Include the MTF as a percentage, the frequency, the field angle, and the wavelength.
Accurate Usage:
The photopic MTF of the lens is 20%@200LP/mm at 50° field angle.
Inaccurate Usage:
The MTF of the lens is 20%.
Instantaneous FoV as Resolution
"Angular Resolution" is used by the Aerospace/Defense industries and Computer Vision Engineers
The angular resolution of a system determines the number of pixels per degree of FoV. Or, the number of pixels per unit distance across an object.
Definition:
- The number degrees (or minutes) in object space subtended by a pixel in the image.
How to Calculate Angular Resolution:
Use Our Advanced Field of View Calculator which includes angular resolution for distortion profiles.
How to Express Angular Resolution:
Include both the angular resolution and the field angle.
Accurate Usage:
The imaging system has an angular resolution of 12px per degree at 25° field angle.
Inaccurate Usage:
The imaging system has an angular resolution of 12 pixels per degree.
What is The Impact of Angular Resolution on Computer Vision?
Angular resolution defines the largest object scale that can be input into a network. The impact on performance is direct. Dollar et. Al shows the direct result on average miss rate of pedestrian detectors. With less than 50 pixels, even state of the art detectors can have a 50% miss rate on detection. This increases rapidly as scale decreases.
From: Dollár, et. al. “Pedestrian Detection: An Evaluation of the State of the Art”
Combined Megapixels and MTF as Resolution
"Effective Resolution" is used by Lens Supplier marketing teams
The effective resolution is a quick method to filter down a lens selection when you are creating a new camera. The optical resolution (MTF) of lenses can be challenging to understand without extensive experience in camera hardware. Many experienced camera engineers do even not know the tricks which manufacturers can play with MTF charts/testing, so effective resolution can be a shortcut directly to system-level resolution.
Definition:
- The sensor pixel count resolution for which a lens' optical resolution is supposed to be suitable.
How to Express the Effective Resolution:
Include both the Pixel Count and the image sensor format size.
Accurate Usage:
The lens is suitable for a 12MP 1/2.3" sensor.
Inaccurate Usage:
The lens is rated for 12MP so it's optical resolution is suitable for any 12MP sensor.
This misleading metric will forever remain used and abused by lens manufacturers. If you are a lens buyer: you have now been warned.
Reference and related links:
1.) Dollár, et. al. “Pedestrian Detection: An Evaluation of the State of the Art”
2.) Vasiljevic, et. Al. “Examining the Impact of Blur on Recognition by Convolutional Networks”
So, How do we at Commonlands use resolution when talking about our lenses?
As a lens supplier, we provide an objective Effective Resolution score based on empirical System Resolution measurements.
What matters for performance is the final image quality.
For general lens effective resolution metrics:
We perform through-focus ISO12233:2014 measurements using a number of sensors that have different pixel pitch.
Then, we find the maximum image circle.
After applying the maximum sensor size and aspect ratio, we calculate the total number of pixels.
For effective resolution metrics of each Sensor Format Type:
We take the same pixel pitch measurement from the general lens effective resolution.
Then, we calculate the number of pixels that would fit on a 4:3 aspect ratio sensor of the specific Format Type.
View Other Image Quality and Computer Vision Topics
- Camera Exposure and Computer Vision
- Motion Blur
- High Dynamic Range
- Resolution and Sharpness
- Shading and Vignetting
- Noise
- Fisheye Distortion and Wide Angle Lenses
- Fringing and Chromatic Aberration
Trying to Determine Your Camera Requirements?
Use our free web-based AoV Calculator to determine your system's Field of View Requirements. Then, use the M12 Lens calculator to match your requirements with the available lenses. Our Depth of Field Calculator also provides the hyperfocal distance and depth of field for every sensor and lens combination.
We also have a couple of other calculators that many engineers find interesting.
