Improving the restoration of textured objects with prefiltering

The restoration of images acquired by a fluorescence microscope is in the presence of noise known to be a difficult problem. This presence of noise can result in restored images that contain severe noise artifacts, or that are smoothed by a strongly imposed regularization. In both cases, small textures present in the image, are lost in the restoration procedure. In this paper we propose a prefiltering method to reduce the noise in the image without hampering the object. We have applied a Gaussian filter and a median filter prior to the restoration, and show how we can compensate for the extra blurring of the Gaussian in the restoration procedure. We show how this prefiltering improves the performance of the restoration algorithms on textured objects. The experiments were performed on textured disks convolved with the point spread function of a fluorescence microscope and distorted with Poisson noise.