Accurate and timely prediction of crop growth is of great significance to ensure crop yields and researchers have developed several crop models for the prediction of crop growth. However, there are large difference between the simulation results obtained by the crop models and the actual results, thus in this paper, we proposed to combine the simulation results with the collected crop data for data assimilation so that the accuracy of prediction will be improved. In this paper, an EnKF-LSTM data assimilation method for various crops is proposed by combining ensemble Kalman filter and LSTM neural network, which effectively avoids the overfitting problem of existing data assimilation methods and eliminates the uncertainty of the measured data. The verification of the proposed EnKF-LSTM method and the comparison of the proposed method with other data assimilation methods were performed using datasets collected by sensor equipment deployed on a farm.