Surround-View Fisheye Optics in Computer Vision and Simulation: Survey and Challenges

In this paper, we provide a survey on automotive surround-view fisheyeoptics, with an emphasis on the impact of optical artifacts on computer visiontasks in autonomous driving and ADAS. The automotive industry has advanced inapplying state-of-the-art computer vision to enhance road safety and provideautomated driving functionality. When using camera systems on vehicles, thereis a particular need for a wide field of view to capture the entire vehicle'ssurroundings, in areas such as low-speed maneuvering, automated parking, andcocoon sensing. However, one crucial challenge in surround-view cameras is thestrong optical aberrations of the fisheye camera, which is an area that hasreceived little attention in the literature. Additionally, a comprehensivedataset is needed for testing safety-critical scenarios in vehicle automation.The industry has turned to simulation as a cost-effective strategy for creatingsynthetic datasets with surround-view camera imagery. We examine differentsimulation methods (such as model-driven and data-driven simulations) anddiscuss the simulators' ability (or lack thereof) to model real-world opticalperformance. Overall, this paper highlights the optical aberrations inautomotive fisheye datasets, and the limitations of optical reality insimulated fisheye datasets, with a focus on computer vision in surround-viewoptical systems.