Kalman algorithm is widely used in the fields of target tracking, wind direction prediction and nonuniformity correction. The prediction of scanning gray value of a linear array camera by Kalman algorithm is proposed. According to the real-time scanning of the linear array detector, the change in the gray level histogram is derived and normalized into Kalman''s formula. The statistical distribution of next time gray level is measured quickly by a recursive method. The integration time is predicted indirectly. Finally, the adaptive control of the dynamic range of the linear infrared detector is achieved. Several 8 bit images are used to provide simulation environment for different scenes. Then, the gray level distribution statistics in neat stage is predicted quickly and stably. The result provides an intelligent solution for the scanning of the linear array infrared camera.