In this demo, we will demonstrate how cameras equipped with super-wide-angle lens and sensor can be specified, simulated, and designed with the purpose of improving machine perception.

We will also show how a specific pixel processing called adaptive dewarping can increase machine perception’s accuracy.

For the demo, we will make use of 2 applications scenario: monocular single frame depth perception and object classification, such as YoloV4.