Design a Morphological De-ringing Filter of Ultrasound Images

In recent years, still image compression technique plays an important role in many applications such as data transmission and data storage. For speeding up image propagation and reducing image storage efficiently, there are many image compression techniques were proposed. Among them, JPEG2000 is a new generation technique, which can encode images at very low bit-rate with acceptable quality. Because its coder is based on wavelet transform, there will be some ringing artifacts in the reconstructed image. In order to improve the image quality, some de-ringing algorithm must be applied to it. In this paper, a set of morphological filters to reduce the ringing artifacts of ultrasound images is presented. It uses 8 predefined morphological operations, including 4 structuring elements (SE) with both dilation and erosion. By the voting strategy, the most suitable morphological filter for each block is selected. To decrease the filter selection information, the side information must be further compressed without any loss. We can get a more acceptable image by applying the proper morphological filter in each block. Experimental results show that the proposed technique improves the quality of the reconstructed ultrasound image in both PSNR and perceptual result when compare to JPEG2000 at the same bit rate.