Abstract: Sparse traffic trajectory data can be used to detect fake plate vehicles effectively. Most existing techniques are based on rules to search for fake plate vehicles records, but they generally don't provide intuitive meanings of rules and their impacts on results. Moreover, they provides less analysis on the behavior patterns of fake plate vehicles. This paper presents a visual analysis system to interactively detect and analyze fake plate vehicles based on sparse traffic trajectory data. We first design a visual query model based on the rule-based fake plate vehicles detection algorithm, then use maps and timelines to display temporal spatial distribution of different types of fake plate vehicles, especially for cell locations and times that fake plate vehicles appear multiple times. The statistical information on a single cell and a cell pair is presented by Cir Flow and histograms, to visually analyze the impacts of different rules on the results. Finally, Case studies show the effectiveness of our system.