An autonomous vehicle is a comprehensive intelligent system that includes environment sensing, vehicle localization, path planning and decision-making control, of which environment sensing technology is a prerequisite for realizing autonomous driving. In the early days, vehicles sensed the surrounding environment through sensors such as cameras, radar, and lidar. With the development of 5G technology and the Vehicle-to-everything (V2X), other information from the roadside can also be received by vehicles. Such as traffic jam ahead, construction road occupation, school area, current traffic density, crowd density, etc. Such information can help the autonomous driving system understand the current driving environment more clearly. Vehicles are no longer limited to areas that can be sensed by sensors. Vehicles with different autonomous driving levels have different adaptability to the environment. If the current driving environment can be evaluated based on the above information, it can help the vehicle determine whether it needs to exit autonomous driving state. Road scene complexity evaluation is a comprehensive result that needs to consider all factors that may affect vehicle driving. Currently, there are no internationally harmonized evaluation methods or indicators. Many scholars determine scene complexity based on mathematical models or subjective evaluations. This paper proposes a neural network-based road scene complexity assessment method to obtain stable assessment results through data drive. We disassemble the scene elements into road elements, human elements, and traffic flow elements, and use these elements as inputs to a neural network that outputs a series of quantities related to the complexity of the scene. Finally, these outputs are weighted and summed to obtain the final road scene complexity score. The method proposed in this paper can largely get rid of the subjective evaluation of scene complexity, and with this method, we can add arbitrary elements (e.g., weather elements, electromagnetic environment elements) to make an evaluation of scene complexity.