Based on the application of software model building, virtual experiment can be conducted with vehicle model, road environment model and sensor model building to develop camera-sensor-based lane detection system in PreScan. These three types of models are connected and the virtual experimental environment is built. Sensorextracted data are processed by Matlab/Simulink to acquire the accurate ground truth. Meanwhile, images from camera sensor are processed with standard size for further application. The virtual experimental environment design is the foundation for the analysis and comparison of ground true and camera-detected data, as well as for accuracy evaluation in system pre-development. In conclusion, the virtual experimental environment design can be used in camera-sensor-based lane detection system development and is a supplement to traditional system development.