In order to find out the high correlative factors of the crop disease from the images, and to predict the disease process, the author analyze the digital images of cucumber leaf which collected in the process of normal and disease by computer image processing technology. By comparing the nine parameters of the three kinds of color spaces which were in common use, it was found that three of them have significant differences. In the further statistic study, the optimal distinguish intervals of

more »

... three parameters were found. All of the distinguish rates of G (green), V (brightness) and Cr (red with the difference between the reference value) were more than 80%, and the optimal distinguish interval of them was 149bit, 152bit, 110bit, which can be used as the feature parameters in predicting the process of cucumber crop disease, and become an important reference in the establishment of the prediction model. The test results show that it is feasible to combined predicting cucumber crop disease process by the color analysis of the digital images with showing the process dynamically in display devices by virtual reality technology, which enables the viewer to observe the overall disease process of the crops, and to have a more intuitive feeling of the pathological changes at different time.