dots_image

dots_image ( Image : DotImage : Diameter, FilterType, PixelShift : )

Enhance circular dots in an image.

dots_image enhances circular dots of diameter Diameter in the input image Image. Hence, dots_image is especially suited for the segmentation of dot prints, e.g., in OCR applications. The enhancement is performed by using matched filters with filter masks that are tuned for a particular dot size. For example, for Diameter = 5 the filter mask is given by:

      |         -21 -21 -21         |
      |     -21  16  16  16 -21     |
 1    | -21  16  16  16  16  16 -21 |
--- * | -21  16  16  16  16  16 -21 |
336   | -21  16  16  16  16  16 -21 |
      |     -21  16  16  16 -21     |
      |         -21 -21 -21         |

The parameter FilterType selects whether dark, light, or all dots in the image should be enhanced. The PixelShift can be used either to increase the contrast of the output image (PixelShift > 0) or to dampen the values in extremly bright areas that would be cut off otherwise (PixelShift = -1).

Parameters

`Image` (input_object)	(multichannel-)image(-array) `->` object : byte / uint2
Input image.

`DotImage` (output_object)	(multichannel-)image(-array) `->` object : byte / uint2
Output image.

`Diameter` (input_control)	integer `->` integer
Diameter of the dots to be enhanced.
Default value: 5
List of values: 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23

`FilterType` (input_control)	string `->` string
Enhance dark, light, or all dots.
Default value: 'light'
List of values: 'dark', 'light', 'all'

`PixelShift` (input_control)	integer `->` integer
Shift of the filter response.
Default value: 0
List of values: -1, 0, 1, 2

Parallelization Information

dots_image is reentrant and automatically parallelized (on tuple level, channel level, domain level).

Possible Successors

threshold

Alternatives

laplace, laplace_of_gauss, diff_of_gauss, derivate_gauss, convol_image

Module

Foundation