4.3 Forest Road Detection

4.3 Forest Road Detection
Since we have derived the Digital Terrain Model of the study area we must detect the forest roads. The centerline of each forest road is required in order to derive road profiles based on which the road technical assessment will be done. For the extraction of the road centerlines we need to detect where exist forest roads on the DTM. More specifically, it is needed that the forest road borders are digitized so that these roads will be separated from the rest terrain. The available input data is the DTM of the study area. The elevation differences between the forest road locations and the rest terrain are small. The borders between the forest roads and the rest terrain are not sharp. Therefore, the fuzzy logic membership
In order to determine the mean elevation of the pixels that belong to the roads the following process has been followed. During the LiDAR data collection the route was on the roads surface. Therefore, the trajectory line of the ZEB1 LiDAR can be used. This is very important information because it allows us to find points that are located on the roads. The LiDAR point cloud is already classified between terrain and non terrain points. Now we are interested to the terrain points. The trajectory line is divided into equal intervals every 2 m and the ground points that are included into a ring of 1 m radius are selected. It is assumed that the points that belong to the same neighborhood have similar elevation. The mean elevation of the points included into each of these rings every 2 meters along the trajectory is calculated, as well as the standard deviation. These particular elevation means and standard deviations are calculated for all the rings on every road. The values of average elevation and standard deviation refer to points which are located definitely on the roads, and can be used in order to determine the mean and standard deviation parameters of the Gaussian curve for each ring. The fuzzy logic Gaussian membership function is applied to DTM segments of 2 meters length and the forest road is detected locally. The fuzzy result that is received
Of course these technical parameters are not sufficient for a complete technical assessment study, but they are the cornerstone of it. The detection of these characteristics is one of the main objectives of this study. The borders of the roads have been determined in the forest road detection part. Roadway width is defined by these forest road borders. Road centerline is needed for the calculation of lengthwise inclinations. This centerline is derived by collapsing the road borders in ArcGIS, while some manual corrections at specific locations are needed. The process that has been followed for the technical assessment of the roads regarding the above mentioned parameters is the