Mathematical basics and related operations of Digital Image Processing: vertex operations

Mathematical basics and related operations of Digital Image Processing: vertex operations

(1) linear vertex operation:

> Mission ();

> Aw.imread('guet.gif ');

Figure (1); imshow ();

B1 = a + 45;

Figure (2); imshow (B1 );

> B2= 1.2 *;

> Figure (3); imshow (B2)

> B3= 0.65 *;

> Figure (4); imshow (B3 );

>>B4 =-double (A) + 255;

> Figure (5); imshow (B4 );

> B4 =-double ();

> Figure (5); imshow (B4 );

> B4 =-double (A) + 100; % B4 = (-1) * A + 100; perform image population. Convert the type to double.

> Figure (5); imshow (B4 );

> B5 = (-1) * A + 100;

> Figure (6); imshow (B5 );

> B65 = (-0.5) * A + 100;

> Figure (7); imshow (b65 );

> B65 = (-2) * A + 100;

> Figure (7); imshow (b65 );

> B65 =-double (A) + 100;

> Figure (7); imshow (b65 );

Conclusion: When the gray-scale transformation of the GST function is linear, that is, B = f (a) = AA + P, the transformation is called a linear vertex operation.

Apparently: When a = 1, p! When the value is 0, the gray level of the image is increased or decreased;

When a> 1, the contrast increases;

When 0 <A <1, the contrast is reduced;

When a is less than 0, the bright area of the image is dimmed, and the dark area is highlighted, that is, the image is supplemented.

(2) Non-linear vertex operation: displays the distortion of the image relative to the original image, eliminates the distortion caused by the dynamic range, and uses the non-linear vertex operation.

There are several transformation formulas:

Round (B) = a + A * a * (max (a)-)

Example: round (x) = (x) + 0.06 * X. * (255-x)

> Aw.imread('guet.gif ');

> Figure (1); imshow ();

> Xlabel ('original image ');

>

> X =;

> Y = x + X. * (255-x)/255;

> Figure (2); imshow (y );

> Figure (3); plot (x, y );

> Xlabel ('graph of the B function ');

> B1 = double (A) + 0.06 * double (a). * (255-double ());

> Figure (4); imshow (B1 );

> Figure (5); imshow (uint8 (B1 ));

(3) histogram changes:

Linear transformation only translates the histogram of the image after it is stretched, and the shape is basically not changed. parts beyond [0,255] are accumulated on the boundary.

> Clear

> Aw.imread('guet.gif ');

> Imhist ();

> Help hist

> Title ('histogram of the original image ');

> B1 = 1.25 * double (A) + 45;

> Figure (2); imhist (uint8 (B1 ));

> Title ('histogram after Transformation ');

(4) histogram equalization:

Convert the input image to a gray level with the same pixel (even if the output histogram is flat, the distribution is even) to enhance the overall contrast of the image.

Clear

Histgram = zeros (1,256 );

CDF = zeros (1,256 );

Export cm,map]=imread('guet.gif ');

[A, B] = size (CM );

For I = 1:

For j = 1: B

K = cm (I, j );

Histgram (K) = histgram (k) + 1;

End

End

CDF (1) = histgram (1 );

For I = 2:256

CDF (I) = CDF (I-1) + histgram (I );

End

For I = 1:

For j = 1: B

K = cm (I, j );

Cm_equ (I, j) = CDF (k) * 256/(A * B );

End

End

Imshow (uint8 (cm_equ ));

Figure (2); imhist (uint8 (cm_equ ));