Skip to content

Home

Jump to bottom
Brian Wandell edited this page Sep 10, 2020 · 52 revisions

The Image Systems Engineering Toolbox for Cameras (ISETCAM) is a Matlab toolbox that was developed by ImageVal. The commercial software (ISET) was used by our colleagues in industry as they designed and evaluated imaging systems. ISET was implemented as an integrated suite of Matlab functions and data structures that

  • Represent quantitative radiometric properties of scenes and illuminants
  • Model image formation through optical elements such as microlens arrays and camera lenses
  • Characterize the sensitivity and noise of sensors, including spatial and spectral sampling of color filter arrays (CFAs)
  • Implement image processing algorithms
  • Calculate a calibrated (radiometric) representation of the rendered image
  • Throughout the simulations real units matter

ISET was (notice past tense) sold through ImageVal, and we shared it (freely) with many academic colleagues, including labs around the world and students who took our Stanford course on Image Systems Engineering. Our colleagues are using the software in innovative ways, integrating ISET with quantitative computer graphics (PBRT), AR/VR design, machine learning applications for rendering (L3) and automotive sensor design.

Today ISET has over 500 registered users who are working in more than 80 companies, 9 research institutes, and 65 universities in 24 different countries. As the usage increases, we believe that keeping ISET commercial would slow progress in developing innovative imaging systems for new applications in consumer, medical and industrial imaging. This perception was reinforced by our experience with the ISETBIO project, an effort that built on ISET to model the biology of the human visual system. Opening ISET to them sped progress.

By creating the ISETCAM repository, which includes the code that pertains to sensor and display models, we hope to support engineering reproducibility, provide educational tools for people entering the field, and speed progress in the exciting field of image systems engineering.

Joyce Farrell and Brian Wandell

Download

As of September 2020 and the Matlab 2020a release, we are dividing the ISETCam downloads into distinct branches. The reason for this is largely because recent Mathworks developments required that we re-write the different windows with their 'appdesigner' framework, replacing the deprecated 'guide' framework. Mathworks has urged us to shift away from 'guide' and we are doing so. However, the 'appdesigner' windows are not supported in 3-year old Matlab versions.

  • The 'stable' branch is based on the guide code and will not be further developed. (Master and stable will be the same, unless we screw up).
  • The 'beta' branch is based on the appdesigner code. It is computationally the same as 'Stable', and it is based on the 2020a appdesigner implementation. We have found that the Mathworks appdesigner implementation has many good features but is not as reliable as we would like.
  • The 'dev' branch is where we do our development. We move the code from this branch is merged into 'Beta', and of course over time 'Beta' will become 'Stable'. We are particularly hopeful that the Mathworks will improve the performance and stability of their opengl and window management.

The core computations are the same in these branches and the differences are mainly confined to the CS framework.

How to cite this work

If you use this toolbox in your published work, please cite this repository and at least one of the articles below. We list them in order of preference:

ISETcam development is led by Brian Wandell's Vistalab group at Stanford University and supported by contributors from other research institutions and industry.

Clone this wiki locally