Many denoising models are available in digital image processing particularly Anisotropic diffsuion model. I am using AD based denoising technique for image smoothing and denoising an image. After denoising an image, what are the factors (for example noise, artifacts, blurr, staircase etc...) we should  notice or quantify on the resultant image to ensure the performance of the denoise. I am not asking about evaluation metrics.