Image Processing Algorithms

This final portion of the simulation is called the image-processing pipeline . In this stage, the image sensor data are converted into RGB display data. ISET includes several algorithms needed for basic camera operation:  exposure duration (auto-exposure), faulty pixel replacement , interpolating missing RGB sensor values (demosaicing), and transforming the sensor color data using methods that account for both the sensor spectral characteristics (color conversion) and the illumination conditions (color-balancing).

Because of the simulator’s extensible and open-source architecture (Matlab), the user can also insert proprietary image-processing algorithms in the pipeline and evaluate how these algorithms will perform under various imaging conditions or with various types of sensors.