The 3D Line Following tool can be used to extract semi-automatically lines directly from one or many point clouds. The tool is semi-automatic because the user must specify a point on the line to extract and a direction to follow. Also note that the tool works whatever the size of the source point clouds are.
The 3D Line Following tool support any type of line that can be isolated spatially (e.g. : transmission line), by intensity (e.g. : road line) or by classification. So in this regard the tool is generic, the only requirement is that the line to
extract can be isolated from other points in the point cloud.
Also the lines to extract don't have to go straight or be on a single plane. It can curve, go up, go down, as long as there is no abrupt direction change, the 3D Line Following tool should be able to extract it.
Performance Tip : If you want to speed up the extraction ensure to exclude as many unnecessary points as possible. For example, if you want to extract a transmission line surrounded by trees filtering out first the trees (i.e. : using clip, intensity or classification) from the point cloud could speed up significantly the extraction process done afterward.
Here are some videos showing the extraction of two different types of lines :
Curve Road Line
Power Line
The 3D Line Following tool has been designed to be easy to use, so that only a few simple parameters
need to be set in order to get great extraction performance.
Probably the most important parameter, the maximum line width tells the tool the maximum width of the line to extract. Note that it is usually better to specify a maximum line width a bit bigger than the width of the line to extract.
Performance Tip : For very sparse line whose width is represented by a single points (i.e. : the width of the line cannot be measured) use a maximum line width that is bigger than the presumed width of the line, but still smaller than the clearance corridor (i.e. : region around the line where no other points is present) around the line.
Performance Tip : Using a very small maximum line width can significantly decrease the speed extraction and the possibly the ability to follow the line up to its end.
Performance Tip : Using a very large maximum line width can possibly decrease a bit the precision of the extracted line.
The fixed segment length let you specify the distance between vertices of the DGN line created by the tool. The exception is that the distance between the ending vertex of the line extracted by the tool and the preceding vertex can be smaller than the fixed segment length specified. Note that this setting is optional. If not specify, the vertices of the created DGN line will correspond to the vertices found by the 3D following algorithm.
Performance Tip : Using a very small fixed segment length for a very long line can increase significantly the time of the extraction and the memory taken by the resulting line. As a rule of thumb a line should not have more than a few thousand vertices.
Performance Tip : The fixed segment length processing is done after the line extraction process, so it is impossible to improve the precision of the line extracted using this setting. On the contrary, using a too big fixed segment length will decrease the precision of the created DGN line.
The intersection setting allows you to activate the intersection detection mechanism which, as the name implied, detects line intersection. When such an intersection is detected the tool stops the line extraction and gives back the control to the user. At that point the user can tell the tool to continue the extraction along one of the segments connecting the intersection or stop the extraction.
Performance Tip : The intersection mechanism takes a significant amount of processing power so it is recommended to activate it only when required.
Here is a video showing how the intersection detection works :
One of the best way to isolate lines from a point cloud is to use the amount of reflecting light (i.e. : intensity) of individual points. The intensity settings can be used to easily determine the intensity range of the line to extract. Note that the tool works also with point cloud having no intensity information. In that case all points of the point cloud are considered and the points of such point cloud are never highlighted by the intensity range picker.
Performance Tip : It is always good to use the narrowest intensity range possible to avoid diminishing
the speed or extraction performance of the tool.
The gap jumping functionality has two main purposes :
The maximum length and search angle parameters are used to determine the 3D cone used as the search area. Note that the search is done from the tip of the cone up to the bottom of the cone or until some points are found.
Here is a video showing how the gap jumping works :