Ptychography is a diffraction-based x-ray microscopy method that removes the resolution limit imposed by image-forming optical elements. However, background noise in the recorded diffraction patterns will degrade the reconstructed images and may cause reconstruction failure. Removal of the background noise from a ptychography dataset is an important but rather ambiguous prereconstruction data processing step because high-spatial-frequency diffraction signals are inevitably partly wiped out along with the noise. In this paper, several newly designed techniques for removing background noise from experimental ptychographic datasets are provided. Meanwhile, effects of residual background noise and high-frequency signal loss on reconstructed image quality are discussed in detail. The image quality is assessed quantitatively by the power spectral density analysis method and spatial resolution calculation. Both the simulated and experimental results indicate that the positive effect of noise removal by these methods clearly exceeds the negative effect of the accompanied high-spatial-frequency signal loss because part of the lost signals can be recovered by the improved consistencies between neighboring diffraction patterns by the noise removal.