Automotive imaging sensor cameras are important parts of ADAS and autonomous driving systems. They provide visual information about car surround environment including shape, color, and texture of different objects such as vulnerable road users, traffic signals, signs, cars, trucks, road markings, and so on. We presented comparative analysis of modern high dynamic range (HDR) automotive image sensors. These sensors included sequential exposure, split pixel (SP), and super-exposure pixel (SE) architectures. Sequential exposure dual conversion gain (DCG) pixels are well known but suffer from motion and LED flicker artefacts. Both SP DCG and SE DCG architectures showed significant improvement of motion artefacts and LED flicker. Focusing on signal to noise (SNR) and resolution characteristics of these sensors we studied object detection capabilities of these sensors utilizing an SNR of ideal observer (SNRI) probabilistic metric. We showed that the SE pixels outperformed the SP pixels capabilities across the automotive temperatures. We also showed that in typical automotive camera thermal conditions smaller 2.1 µm pixels may additionally provide for better detection capabilities in comparison to larger 3 µm pixels. The super-exposure pixel sensors provide the best foundation for object detection, especially in corner cases where extra SNR and resolution at higher temperatures improve algorithm performance. ‘This improved object detection performance enables objects to be detected or classified at longer distances, giving the ADAS system in the vehicle more time to analyse, plan and act, which will improve safety metrics at a constant speed, or enable a fixed Use Case at higher speeds, ex. more freeway autonomy or automated driving above standard speed limits.