Noise estimation and processing for SLM microscopic images

Because of the characteristics of stereo light microscope (SLM) micro stereovision system, noise is unavoidable in SLM microscopic images. It may have adverse effects on many techniques relative to images. The noise is analysed based on micro stereovision system and divided into random noise and fixed pattern noise, and then an image model is developed. The random noises in SLM microscopic images with various magnifications are quantified and the tiny increasing trend of random noise with the magnification becoming higher is found. Moreover, an extended vector median filter and an average calibration method are used to remove random noise and non-uniformity correction. The experiment results show that the random noise and fixed pattern non-uniformity can be removed efficiently with non-uniformity correction and vector median filtering.