Abstract:Target reconstruction and separation from background under low SNR are key issues in the current image reconstruction task. Photon waveform correction is an effective method to reduce photon accumulation, but it has the problems of tail noise accumulation and long running time when reconstructing at a low frame rate. The runtime is positively correlated with the length of the gating acquisition time channel, which makes it difficult to meet the needs of real-time reconstruction. Based on the formula of waveform correction and the uniform distribution of background noise in the time domain, the paper proposed a signal-filtering algorithm based on noise calculation and an image-filtering algorithm based on region growth. The experimental results show that compared to the waveform correction method, the proposed signal filtering algorithm in this paper has a shorter running time, and achieves a target recovery degree that is more than 1.2 times that of the matching filtering method. The target recovery degree of the image filtering algorithm is increased to more than 1.4 times compared to the variance feature separation. The combination of a signal filtering algorithm and an image filtering algorithm improves the separation effect between the target and the background in the aerospace background.