This paper proposes a denoising algorithm for physical, electronic images based on fractional differential equations. The algorithm effectively combines fractional calculus theory and gradient descent flow. At the same time, we introduce the time factor into the improved denoising model based on the spatial fractional partial differential equation. We take advantage of the unique amplitude-frequency characteristic of fractional differential operation to preserve the texture details with little grayscale change in the smooth area of the image. The model realizes the simultaneous denoising of physical, electronic images in the time direction and the spaceplane. The experimental results show that the fractional-order partial differential equation method has more advantages than the integer-order partial differential equation in denoising and reducing the staircase effect.