Morphometric criteria for evaluating individual territorial units in solving the problem of spatial development of linear objects of terrestrial transport infrastructure

Background. The article discusses the issue of evaluating the morphometric properties of a polygon in order to form a rating matrix necessary to solve the problem of finding a path in a discrete space.  This approach to solving the problem of spatial development of linear objects of land transport will be useful in solving practical problems of camera tracing.

Materials and methods. The landfill in question is separated into separate territorial units, each of which has its own rating based on a set of different criteria (anthropogenic load, morphometry, etc.). The rating of a particular territorial unit determines the prospects of its consideration when solving the tracing problem. Such a discrete polygon model is actually a weighted graph, in which the search for a path is carried out by various search algorithms: A*, Dijkstra, Breadth-first search. Since the assessment of a particular territorial unit based on the morphometric properties of an area depends on its size, it is proposed to evaluate the polygon in two stages with different resolutions of discrete space, i.e. enlarging individual territorial units into blocks. As an example, in the work, the initial high-resolution height matrix was programmatically generated, which was later combined into blocks.  With respect to the matrix of formed blocks, sensitivity analysis was performed according to 4 criteria: the average slope of the block, the standard deviation of heights, the height range and the roughness index.

Results. It follows from the experimental results that the "Height range" criterion has the highest average sensitivity and, as a result, weight. This means that this criterion has a greater impact on the formation of block ratings. At the same time, the graph of rating changes according to this criterion and the color matrix of the weight range allow us to conclude that this criterion is the least informative, since it weakly distinguishes the blocks from each other.