Abstarct: One of the main topics in image processing is salt and pepper impulsive noise removal from digital images. This type of noise randomly generating white (salt) and black (pepper) points in images and causing the loss of information and details in images. The purpose of this paper is designing an algorithm to remove impulsive noise from gray and color images using combination of median and mean filters. In addition to high efficiency in noise removal, volume calculations are not complicated. The proposed method consists of two phases identify noise and noise removal. In first considered the low and high threshold values to identify noisy pixels. In the second stage the decision is using the adaptive median and mean filters baxsed on the 1st order neighborhood and diagonal neighborhood pixels. Furthermore, the proposed algorithm is compared with recent articles using PSNR, MSE, MAE and SSIM parameters in Different densities impulsive noise. Visual and numerical results show good performance of the proposed method is the removal of impulsive noise. In the end, according to parallel processing and real-time features in FPGA chip, proposed method using VHDL hardware descxription language writing, synthesized and implemented on Xilinx Spartan-6 series XC3SLX9. The total power consumption, maximum output frequency and average processing time in FPGA for a color image with dimensions 256×256 and 90% densities impulsive noise, respectively values of 14mw, 42.186 MHz and 1.299 sec is obtained Which reflects the performance of the proposed system is efficient and fast.