next up previous contents
Next: Complex Images Up: Image Processing Previous: Convolution and Filtering   Contents

Pixel Processing

General pixel based manipulation including transendental functions is carried out by the following routines. For multiple image manipulation the image type returned is determined by the function im_sup_vtype

Vartype im_sup_vtype(Vartype vtype1, Vartype vtype2)
which ensures no truncation of the result due to casting (truncation may still occur for numbers beyond the normal range of number representation).

Imrect *im_add(double k,Imrect *im)
Add the constant k to all pixels in the specified image. For the case of complex images only the real component is modified.

Imrect *im_diff(Imrect * im1, Imrect * im2)
Subtract the pixels contained in two images and return the difference image.

Imrect *im_minus(Imrect * im)
Inverts the sign of all pixel values returning a new image.

void    im_pixf_dec(Imrect * image, int i, int j)
Decrement pixel (i, j) in image. Pixels outside image are ignored.

Imrect *im_sum(Imrect * im1, Imrect * im2)
Returns an image whose pixels are the sum of those in the input images within the common region of interest.

Imrect *im_times(double k, Imrect * im)
Returns an image with pixels scaled by 6#6 of their input values.

Imrect *im_prod(Imrect * im1, Imrect * im2)
Returns an image with values given by the product of the pixels in the images im1 and im2.

Imrect *im_div(Imrect * im1, Imrect * im2, double thresh, double val)
Divide the pixels contained in image im1 by those contained in image im2. Numerical stability is maintained by preventing the denominator reducing below a value of 7#7 when it's absolute value is less than 8#8.

Imrect *im_sqr(Imrect * im)
Compute the sqaure of each image pixel value, all values returned as float_v except complex_v, which is returned as complex_v.

void    im_grad(Imrect * im, Imrect ** imx, Imrect ** imy)
Compute derivative images for both the x and y components using the functions imf_diffx(im) and imf_diffx(im).

void    im_hessian(Imrect * im, Imrect ** imx, Imrect ** imy, Imrect ** imxx, Im
rect ** imxy, Imrect ** imyy)
Computes both x and y derivatives, and second derivatives. If the first derivatives are not required then pass null pointers for imx and imy.

Imrect *im_log(Imrect * im)
Returns an image of natural logariths for all pixels while maintaining sign. The function is a direct inverse of im_exp(). All images are returned as type float_v except for type complex_v which is returned as complex_v.

Imrect *im_quad(im)
Returns an image which has been vertically and horizontally doubled to produce boudary continuity. Intended for use before Fourier transform operations to remove Gibbs oscillations after deconvolution.

Imrect *im_sin(Imrect * im)
Returns an image with each pixel given by the trigonometric sine of the initial grey level data.

Imrect *im_sqrt(Imrect * im)
Returns an image with each pixel the square-root of the initial grey level value. The sign of the original pixel is preserved, for true complex treatment of negative values use imz_sqrt().

Imrect *im_asin(Imrect * im)
Computes the trigonometric sine of all pixels in the image. All returned images are of type float_v except complex images which are returned as type complex_v.

Imrect *im_exp(Imrect * im)
Compute the exponentiated value of each pixel with maintained sign. Returns images of type float_v for all cases except complex_v which are returned as complex_v. The function is the direct inverse of im_log().

void im_scale_range_inplace(Imrect * im,
                            double oldlow, double oldhigh,
                            double newlow, double newhigh,
                            double threslow, double threshigh)
This routine scales the range oldlow, oldhigh in image im newlow, newhigh, thresholding at the upper and lower values given by threslow, threshigh.

void im_gamma_scale_range_inplace(Imrect * im,
                                  double gamma,
                                  double oldlow, double oldhigh,
                                  double newlow, double newhigh,
                                  double threslow, double threshigh)
Gamma scale the range oldlow, oldhigh in image im using the specified parameters.

Imrect *im_shading(Imrect *im,double slant,double tilt,double scale)
Produces a shaded image from a 3D image surface using a point light source and lambertian reflectance model at a distant location specified by 9#9 and 10#10 .

Imrect *im_combine(Imrect * im1, Imrect * im2, void *(*func) (), void *data)
Returns an image with pixels computed from the specified input image pixels operated upon by the specified function.

Imrect *im_fpp_combine(Imrect * im1, Imrect * im2, void *(*func) (), void *data)
General purpose routine for operating on two images of types float_v and ptr_v with the specified function.


next up previous contents
Next: Complex Images Up: Image Processing Previous: Convolution and Filtering   Contents
root 2017-11-24