Performs anisotropic diffusion on an image

dst = cv.anisotropicDiffusion(src)
dst = cv.anisotropicDiffusion(src, 'OptionName',optionValue, ...)

Input

• src 8-bit 3-channel source image.

Output

• dst Destination image of the same size and the same number of channels as src.

Options

• Alpha The amount of time to step forward by on each iteration (normally, it's between 0 and 1). default 1.0
• K sensitivity to the edges. default 0.02
• Iterations The number of iterations. default 10

The function applies Perona-Malik anisotropic diffusion to an image. This is the solution to the partial differential equation:

dI/dt = div(c(x,y,t) * nabla_I) = nabla_c * nabla_I + c(x,y,t) * delta_I

where delta denotes the Laplacian, nabla denotes the gradient, div is the divergence operator and c(x,y,t) is the diffusion coefficient (usually chosen as a function of the image gradient so as to preserve edges in the image).

Suggested functions for c(x,y,t) are:

• Exponential Flux: c(||nabla_I||) = exp(-(||nabla_I||/K)^2)
• Inverse Quadratic Flux: c(||nabla_I||) = 1 / (1 + (||nabla_I||/K)^2)

OpenCV implements the first option. The constant K controls the sensitivity to edges.

References

[PeronaMalik90]: P. Perona and J. Malik, "Scale-space and edge detection using anisotropic diffusion", in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 12, no. 7, pp. 629-639, Jul 1990.