A Real Time Speckle Noise Cleaning Filter for Ultrasound Images

Ultrasound images are masked by multiplicative speckle noise caused by random interference between coherent backscattered waves. This paper presents an efficient algorithm for real-time speckle cleaning filter of ultrasound images and its VLSI implementation in FPGA. The algorithm presented here implements a regular, tunable filter with an optimized architecture having proper parallelism and pipelining to achieve high speed. The performance of the proposed filter is compared to that of the aggressive region growing filter (ARGF). The processed images obtained from the proposed filter are speckle removed and the edge details are also restored effectively. Also, the PSNR performance is better than the ARGF filter while noise variance is lower indicating a cleaner image as more speckle is reduced