Canny Filter Application Implementation with GRASS

In this work an edge detecting filter using the Canny algorythm is implemented and applied to test images in order to assess its potential in detecting borders of elements. The filter will also be used on a real application in order to detect changes in image geometry. The role played by GRASS GIS is crucial because of its open source written in C language, its portability on different platforms and the ease of having new code added to implement new modules in a GIS environment. In this publication we present therefore how the implementation of the filter is carried out and how it was used on the test images to check the accuracy of results. Finally it is also applied to a digital image of the side of Stromboli vulcano to check its deformations. From such image the Canny filter can extract the profile of the conical shape of the mountain and compare it in real time with the same profiles taken previously, detecting changes due to erosion or landslides. The result can be sent using internet technology to operators which interpret the activities recorded.