Configure Bitone Edge Contrast Enhancement Parameters dialog
for Image Scanner Implementation Only

TIP: Increasing this value increases the amount of the shadow removed and the size of a white outline to the right and left of dark areas, but may decrease scanning speed. Decreasing this value decreases the amount of the shadow removed and the size of a white outline to the right and left of dark areas, but may increase scanning speed.

A pixel’s contribution to legibility is determined in the context of the surrounding pixels. To provide context, an imaginary box is drawn around suspect pixels ("test pixel") in the image. The width of this imaginary box is Kernel Width, and it equals the number of pixels equally divided half to the left and half to the right of the test pixel. For example, a Kernel Width equal to "7" yields an imaginary box 7 pixels wide; that is, 3 pixels to the left and 3 pixels to the right of the test pixel (3-2-1-Test Pixel -1-2-3).

The value specified must be an odd numeric value.

Range: 3 - 17

Default: 7

TIP: Increasing this value increases the amount of the shadow removed and the size of a white outline above and below the dark areas, but may reduce scanning speed. Decreasing this value decreases the amount of the shadow removed and the size of a white outline above and below the dark areas, but may increase scanning speed.

Kernel Height is similar to Kernel Width. Kernel Height equals the imaginary box’s height in the same manner Kernel Width equals the box’s width. Thus, a Kernel Height = 7 pixels, and a Kernel Width = 7 pixels results in an imaginary box consisting of 49 pixels (7 x 7).

The value specified must be an odd numeric value.

Range: 3 - 17

Default: 7

TIP: Decreasing this value makes the shadow removal or outlining more aggressive at the risk of removing light variable data or other light items on the document. If unacceptable image loss occurs, increasing Kernel Threshold will preserve more light image data.

The grayscale value of every pixel within the imaginary box is compared to the test pixel’s grayscale value. The total number of pixels in the imaginary box that are darker than the test pixel by at least the Kernel Threshold value is counted. Example: consider a Kernel Height = 3 pixels, and a Kernel Width = 3 pixels and a Kernel Threshold = 40. Assume the test pixel’s grayscale value = 251, and the surrounding 8 pixels have grayscale values of 200, 210. 211, 202, 214, 198, 212, 205, respectively. Five (5) pixels are darker (less) than the test pixel by at least the Kernel Threshold value.

Range: 5 - 200

Default: 40

TIP: Decreasing this value makes the shadow removal or outlining more aggressive at the risk of removing light variable data or other light items on the document. If unacceptable image loss occurs, increasing Kernel Threshold Count will preserve more light image data.

Kernel Threshold Count equals the number of pixels surrounding the test pixel that must be darker by at least the Kernel Threshold value to turn the test pixel white. In the example above, if the Kernel Threshold Count = 5, then the test pixel will be made white.

Range: 1 - 289

Default: 5

TIP: Decreasing this value can aid in the removal of very dark shadows or dark shaded backgrounds around text, but may reduce scanning speed. Increasing this value may increase scanning speed, but may inhibit the removal of dark shadows or the removal of dark shaded backgrounds around text.

There exists a pixel grayscale value so dark that it need not be compared against the surrounding pixels in the imaginary box to determine if it should be made white. Pixels whose grayscale values are below the Kernel Keep Black Threshold will be left black without further analysis.

Range: 0 - 255

Default: 100