ppolynomialfitting

Approximates image to flat background using polynomial fitting.

Synopsis

ppolynomialfitting xorder yorder xyorder [im_in|-] [im_mk|-] [im_out|-]

Description

ppolynomialfitting converts an image content into a flat background image, using a polynomial fitting.

The image im_mk is used as a mask, and defines the list of pixels that can be used to compute the polynomial approximation.

The order of the polynomial can be selected separately for x, y, and mixed terms. For example, with orders 2, 3, and 2 for x, y, and xy respectively, the polynomial will be:

  a+b*x+c*x^2+d*y+e*y^2+f*y^3+g*xy

Parameters

xorder is the x order [0..10].
yorder is the y order [0..10].
xyorder is the xy order [0..10].

Inputs

im_in: a 2D float image.
im_mk: a 2D gray image used a mask image.

Outputs

im_out: a 2D image of the same type as the input image.

Result

Returns SUCCESS or FAILURE.

Examples

Correction the illumination of the tangram image using background subtraction. The background is approximated by a polynomial:

   pthresholding 0 73 tangram.pan mask.pan
   ppolynomialfitting 2 2 1 tangram.pan mask.pan a.pan
   pim2sf tangram.pan tangramf.pan
   psub tangramf.pan a.pan b.pan
   pmeanvalue a.pan; mean=`pstatus`
   paddcst $mean b.pan out.pan

More examples

C++ Prototype

Errc PPolynomialFitting( const Imx2d &im_in, const Img2d &im_mk, Imx2d &im_out, int xOrder, int yOrder, int xyOrder );

Version française

Calcul de l'approximation du fond d'une image en utilisant une approximation polynomiale.

Author: Régis Clouard