Basic Description: Edge uses a variety of edge detection algorithms to highlight sharp contrast changes in an image. The resulting image is binary, with edges indicated as value = 1 and all other pixels with value = 0.
Technical Details: Refer to descriptions of each algorithm.
Instructions: After choosing a selection method and making your selection on the survey, choose an algorithm and enter the parameters using the instructions below.
![1](images/drex_module_4_6_20_image_1.png) Canny
Canny is based on the well-known Canny edge detection algorithm. A higher Threshold will return fewer edges at a given scale. Sigma determines the size of the Gaussian filter used to smooth the image before edges are extracted.
|
![2](images/drex_module_4_6_20_image_3.png) Log
LoG using the zero-crossings of the Laplacian of Gaussian filter to detect edges in the selected tiles. A higher Threshold will return fewer edges at a given scale. Sigma determines the size of the Gaussian filter used to smooth the image before edges are extracted.
|
![3](images/drex_module_4_6_20_image_5.png) Prewitt
Implements the standard directional Prewitt edge detection algorithm.
|
![4](images/drex_module_4_6_20_image_7.png) Sobel
Implements the standard directional Sobel edge detection algorithm.
|
![5](images/drex_module_4_6_20_image_9.png) Roberts
Implements the standard Roberts edge detection algorithm.
|
![6](images/drex_module_4_6_20_image_11.png) Zerocross
Zerocross using the zero-crossings of the Laplacian filter to detect edges in the selected tiles. A higher Threshold will return fewer edges. This choice is similar to LoG but does not allow smoothing.
|