The combobox shows all visible layers of the current QGIS project with the geometry type LINESTRING. It is recommended, that this layer should be a clean graph following a node-edge-model, i.e. different lines should only touch other lines, never cross them. You can get souch a clean graph e.g. by running the PostGIS function "CreateTopology".
The combobox shows all visible layers of the current QGIS project with the geometry type POINT.
The combobox shows all fields of the chosen trajectory layer. The selected field will be added to the result features.
Insert the EPSG code, which should be used for the result layer. It is recommended to insert the CRS of the network and trajectory layer (both layer should have the same CRS). It is also recommended to use a metric CRS.
The standard deviation of the distance between the positions from the trajectory and their possible positions on the network. Go to the repository on GitHub for further information about this setting.
The expected distance between the positions from the trajectory and their possible positions on the network. It is recommended to choose the value 0, i.e. ideally, the trajectory points already represent the correct positions.
The maximum distance to find points on the network for each trajectory point. It is madatory, that this value is greater oder equal than the highest shortest distance between each trajectory point and the whole network. This setting has a big influence on the computing time. Go to the repository on GitHub for further information about this setting.
The plugin runs the following tasks:
read the input layers
calculate candidate points for each trajectory point using the maximum search distance and create a graph of candidates
calculate the probability for each candidate point to be emitted by the correpsonding trajectory point
calculate the similarity of the transitions compared to the trajectory points
find the best candidates using Viterbi algorithm
create a layer of linestrings using this candidates
A detailed description about the algorithm and the settings are provided in the README document of the repository on GitHub.