Despeckling of Medical Ultrasound Images Using Kriging Interpolation

The focus of this research paper is to tackle the issue of speckle noise present in medical ultrasound imaging. The reduction of speckle noise is challenging because of its multiplicative nature. To accomplish this task, a new filter is developed based on Kriging interpolation. The proposed filter structure is built upon creating an Euclidean distance matrix based on four statistical features in a confined neighborhood. The center pixel of this neighborhood is replaced with a new value estimated using the weights from the distance matrix. The effectiveness of the proposed filter is tested by applying it to synthetic images and real ultrasound images. Different performance comparisons demonstrate the superiority of the proposed filter over other state-of-the-art methods.