Reflective tomography lidar imaging has the advantage of simple system and is not sensitive to atmospheric turbulence, but ordinary pulse signal is difficult to satisfy the requirement of both high precision and long range. To solve this problem, an imaging processing method based on chirped pulse for reflective tomography lidar is presented. In this method, coherent receiving and compression processing is done for chirped pulse echo firstly. Then, projection data of different observation angle is achieved by extracting envelope. At last, high precision image reconstruction is obtained by using Convolution Back-projection (CBP) algorithm. The results of simulation indicate that the envelope obtained by this method is consistent with reflective projection data at the same projection angle, and the outline image of the target can be obtained when the range of projection angle is larger than 60 degree. Imaging experiment validates the feasibility and effectiveness of the presented method.