Abstract: Based on cultivation experiment with different nitrogenous nutrition levels, the color features of cucumber under varied N-status have been abstracted by the digital camera and image processing, by the method of linear stepwise multiple regression, a prediction model for photosynthetic pigment contents of cucumber leaves at initial bloom stage using color feature parameters was developed, whose precision was evaluated and tested. The results of statistic analysis show that R/(R+G+B) is the best parameter for estimating cucumber leaf chlorophyll content, and G/R and R/(R+G+B) are the best parameters for estimating cucumber leaf carotenoid content. The article may provide a fast method for monitoring the growth of cucumber.