Abstract: In order to achieve rapid non-destructive monitoring of tomato yellow leaf curl virus disease, we analyzed the images of tomato leaves using computer image processing technology. We compared 9 color parameters in 3 color systems. The study showed that there were significant differences in five color parameters. In a further study on distribution statistics, we found the optimal distinguish interval of each parameter. Each distinguish rate of G, Y and Cb is more than 70%. The optimal distinguish interval of each 3 value is 135, 121 and 110. This provided important parameter basis for the follow-up study on identify model. The result shows that it’s feasible to identify tomato yellow leaf curl virus disease based on color analysis.