There have been lots of tools for cognitive neuroscience research and some of them such as eye tracking, EEG grow up quickly. However, as each tool can only complete single function and most of technologies are not very mature, there is still a long road to get a convincing conclusion. In view of that, this paper proposes a tri-modal Schema combining EEG, eye tracking and image recognition for cognitive neuroscience research. The Schema contains the architecture to build such an experiment environment and the steps of data analysis. How to acquiring a signal with few artifacts and synchronize all the signals are focused. It provides a robust solution for establish an experiment environment quickly.