vderiche − Deriche edge operator

vderiche if=infile of=outfile [a=alpha] [w=omega] [-f] [-d]

Vderiche computes the Deriche Edge detection function. All computations are 32 bit floating point. By default the output for floating images is scaled to be in the range 0.0 to 1.0, for integer images the range is 0 to 255. If non-maximum suppression and thresholding with hysteresis is desired use the vedgex(1) command.

Vderiche accepts all numeric VisionX image types. The result type will be the same as the input type or optionally of type float.

a=alpha

The a= parameter specifies a value for alpha, the parameter adjusted to trade- off between signal-to-noise ratio and localization: decreasing it will lower edge localization, but will yield a better SNR, and vice versa. (A large alpha translates to a small ’window’ with good localization. A small alpha means a large ’window’ which smoothes out the image and yields a better SNR). The default value for alpha is 1.0.

w=omega

The w= parameter specifies a value for omega, the second parameter which describes the Deriche operator. As omega becomes smaller, the number of local maxima decreases sensibly. Its default value is 1/1000 of alpha, and it should only be changed for fine-tuning final results.

-d

If the -d option is file with two channels is generated: channel one contains the edge magnitude and channel two contains the angle of the gradient. angle information is scaled to make full use of the 0-255 available range. The four cardinal gradient directions have the following values: East=0, North=64, West=128 and South=192.

-f

If the -f option is selected, the output format is specified to be floating point. (default output format is the same as input format)

D. Griegel, D. Roitman

vedgex(1)