In order to reduce the impact of speckle noise on reconstructed phase quality in the field of digital holography, An improved non local mean filtering method for speckle suppression in digital holography is proposed. This method uses the Pearson correlation coefficient in statistics to improve the measurement method of sub block similarity in non local mean filtering, and divides the similarity coefficient into two different intensity domains based on its size, assigning different values to them, making the contrast of grayscale image denoising stronger. This method can not only further filter out speckle, but also better preserve the edge information of the image, which can solve the problem of image overfiltering in traditional non local mean filtering. The experimental results show that the improved non local mean filtering can effectively reduce speckle noise in digital holographic images, improve image quality and clarity, and perform better than most traditional digital image processing methods.