Design Considerations for Porting Real-Time ADAS Applications to Embedded Platforms

ADAS (Advance driver assistance systems) are used in modern cars until we reach full autonomy. These systems rely on camera to take necessary decision. These applications include features like lane departure warning, surround view, collision warning, object detection, traffic sign recognition and more, all these features require advance machine vision algorithms to perform as per intent. Even though it is easier to develop machine vision algorithm to address these features, porting the same on an embedded platform for real time automotive application is difficult and requires various parameter to be considered.

This case study explains various designs consideration for porting real time ADAS algorithms on a embedded platform like TDA2x and TDA3x SoCs.