Intelligent networked vehicle is a product under the information age, and it is also the main trend of the current automobile development. The purpose of this study is to clarify the adaptive technology of intelligent networked vehicles in complex environments through the use of advanced assisted driving system technology. Target detection, which is one of the basic technologies, is selected as the object of study, and an intelligent algorithm is used to construct a road target detection model. Specifically, the road target detection model based on improved YOLOv5s is constructed by introducing the CBAM attention mechanism, improving the feature fusion structure, integrating the C2f model to optimise feature information extraction, and improving the IOU loss function to optimise the YOLOv5s algorithm. Different datasets are selected for experiments to explore the detection effect of the model on different targets and environments. The results show that the recognition accuracy of this paper’s model in detecting different road targets is above 86%, and the detection accuracy in different environments such as sunny, rainy and nighttime is improved by 3.25% to 4.20% compared with the original YOLOv5s algorithm. By improving the road target detection model of YOLOv5s, the detection accuracy in complex driving scenarios is significantly increased and it better satisfies the requirements of intelligent driving technology for high-performance target detection algorithms.