In order to detect the vehicle stopping in highway tunnel more accurately, the traditional image processing technology is combined with deep learning. Firstly, the foreground moving targets are extracted using the background difference method based on Gaussian Mixture Model (GMM). Then the meanshift algorithm is applied to track these foreground moving targets. By calculating the speed of the moving targets and the correlation of the moving target between the neighbouring video frames, and comparing with the speed threshold and correlation threshold, the static target is detected. Finally, we combine the Convolutional Neural Network (CNN) classification model to identify whether the static target is a vehicle. The method proposed in this work is validated using the real highway tunnel vehicle stopping video and reaches at least 84% accuracy. It is also compared with the traditional image processing method without CNN, which shows that our method improves at least 63% accuracy.