We develop questions around the challenges of automotive image quality and show that especially colour separation probability (CSP) and contrast detection probability (CDP) are key enabler to improve the knowhow and overview of the image quality optimization problem. We investigate image quality challenges and link them to the physical properties of use case scenes. By exploiting the above methods, the new detection probability based KPIs show to be a helpful tool to link image quality to the tasks of autonomous driving. We show the state of the IEEE P2020 discussion around colour separation probability which includes photon shot noise and the metameric properties of light spectra. As a second part we investigate the already presented KPI Contrast Detection Probability and discuss its advantages and limitations towards different metrics of automotive imaging such as HDR, low light performance and object detection.