The detection theory of a photon counting laser range finder system under movement condition is studied. The theoretical issues such as statistic distribution features and detection probability involved with the data processing of the system are analyzed. According to the ranging background under movement condition and the characteristics of echo signals of the system, the simulation signals containing target echo signals and noise echo signals which correspond to the statistic distribution features are generated by Matlab. On the basis of the research on the traditional data processing algorithms such as Poisson filtering algorithm, an improved algorithm based on raster filtering is proposed. The algorithm is used to process the simulated echo wave data. Finally, the noise signals distributing randomly throughout the detection space are filtered out effectively and the motion trajectory of the target is obtained. Compared with the traditional data processing algorithms, this algorithm is more effective in the processing of photon counting laser ranging data under movement condition.