FPGA-based remotely sensed imagery denoising

This paper presents a FPGA (Field Programmable Gate Array)-based remotely sensed imagery denoising method, named FPGA-based median filtering. The proposed method is capable of processing large volume data since it takes full advantages of FPGA hardware and abundant logic units. This paper first overviews the traditional median filtering algorithm, and then highlights the FPGA-based filtering, including hardware components, software platform, and implementation of traditional median filtering using FPGA hardware in detail. The image frame with two dimensions of 10000×10000 pixels² is employed to test the proposed method. The comparison analysis between the proposed FPGA-based median filtering method and traditional median filtering based on ENVI version 4.8 software is carried out. The experimental result demonstrates that the proposed method is capable of saving the time more than 20 times than traditional median filtering based on ENVI version4.8 software does.