Apply the object detection technology to sense the road surface, such as manhole covers and speed bumps ahead the vehicles could adjust the system parameters such as the vehicle chassis suspension height, motion control system steering in advance to improve the comfort and safety of driving. Considering that manhole covers and speed bumps are difficult to be detected accurately under undesirable lighting conditions such as strong daylight and cloudy days, offline data augmentation methods that adjust the saturation, brightness, contrast and sharpness of the image are used to improve the accuracy and robustness of the detection model. The experimental and analytical results show that based on these offline data augmentation methods, the detection accuracy of the model for manhole covers and speed bumps is improved, and the average accuracy reaches 95%, with an improvement of 3.2%.