Edge Detection in Claud Classification

It is a certain number of gradient operators using small arrays of pixels were tested in this study, in order to check their capacity to properly detect significant boundaries between different CTT and eventually cloud types. This capacity is 14 visually tested against direct CTT images. Among the detectors applied are the Canny, Roberts, Sobel and Kayyali 15 SENW (to be named here SENW) edge detectors, and the Harris corner and edge detector. The Roberts edge detector employs mainly 2 pixels in a 2*2 matrix for each computing direction (horizontal and vertical). The test conducted show a lesser sensitivity to edges i.e. the variation range of gradients is limited compared to the other methods.

The Canny 18 method uses a 5*5 pixel matrix, that undergoes preliminary filtering before the use of a smaller-array gradient detector, making the calculation procedure longer. The Sobel detector uses a 3*3 matrix where 6 pixels practically contribute to 20 the gradient. The SENW edge detector is based on the Sobel approach for edge detection but employs mainly 4-corner 21 pixels (upper and lower left, and right) in a 3*3 pixel-matrix, the remaining 5 pixels between the corners are set to zero. The Harris detector detects both edges and corners.