Mobility Map Generation For Ground Surveillance On Hilly Terrain

Saturn Software Mills Visit for more technical article http://www.saturnsoftmills.com/FreeDownloads.htm

Mobility Map Generation for Ground Surveillance on Hilly Terrain ABSTRACT This paper presents a technique to generate mobility map of a particular vehicle for ground surveillance. This technique uses the Digital Terrain Elevation Data (DTED) and the Digital Material Analysis Data (DMAD) as the basic source of information. It also considers the man made (i.e. roads) and natural (i.e. rivers) objects in the calculation of mobility map for a vehicle. This mobility map may provide help to the operation planner to select the best route for vehicle movement over the hilly terrain.

1. Introduction There exist a wide class of problems involving the movement of vehicles on hilly terrain. There are different categories of vehicles, broadly it can be divided in to two categories, wheeled vehicles and tracked vehicles [1]. Tracked vehicles have better cross-country ride and shock performance than wheeled vehicles because of a distinct advantage for tracks in rough terrain. A tracked vehicle can weigh more and carry a larger payload at a greater speed while negotiating soft soil. These vehicles have different climbing angles and varying capability to run on different material type. The operational planner uses the topographic maps and geological maps to analyze & mark the suitable regions for movement of a particular vehicle type. This exercise requires lots of attention and expertise. The accuracy of the result varies depending on the capability of the analyzer. The complete exercise has to be repeated for different vehicle types. The weighted graph technique described in this paper uses the DTED & DMAD as main source of input and generates the mobility map for different vehicles. Once the data

is prepared, it can be used for movement analysis for different vehicles without any repetition. 2. Data Preparation The DTED is the primary input to this technique. It can be generated by various techniques. The digitization of topographic maps generates TIN[2] model that is used to generate the DTED. Sometimes topographic maps are not available of desired area. The SAR interferometery[3] or stereo methods[4] can be used to generate the DTED of desired area. The DMAD can easily be generated by applying classification methods[5] over the satellite image. The information of man made objects like roads and natural objects like river, lakes etc. can be obtained from the topographic maps. This paper presents the results using DTED generated from maps & DMAD data only. 3. Data Representation The ground surface may contain infinite number of small points. This approach divides the surface in to finite number of cells and fits these cells in to a rectangular mesh as shown in figure 1.

Saturn Software Mills Visit for more technical article http://www.saturnsoftmills.com/FreeDownloads.htm

factors that contributes in to the total weight of an edge (say Tw).

Figure 1 : Rectangular Mesh

If S represents the complete surface and it is divided in to 'n' number of equal sized cells C then n

S =

∑ Ci i=1

These cells are base unit of weighted graph technique. A cell represents a uniform area on the surface. The number of cells 'n' to represent a small surface area should be sufficient in number so that the uniformity of the surface can be maintained. Each cell has a corresponding entry in DTED and the DMAD data that will be used in the weighted graph technique to compute the mobility map. 4. Weighted Graph Technique Each cell in rectangular mesh representation of ground surface has eight adjacent cells (excluding boundary cells). All the neighboring cells are connected with eight edges as shown in figure 1. These cells constitute a graph network among all the cells as shown in figure 2. A vehicle has to follow these edges for movement from one cell to another cell. The weight of every edge will be computed by this technique for each edge. This weighted graph results in to a mobility map. There are different

Figure 2 : Graph Network of the ground

4.1 Slope Weight Factor In hilly terrain, the greater slopes are avoided for the movement. Every vehicle has its own capability to climb a particular slope. It is always preferred to follow least slope route for the movement. That is why vehicle follows circular route while climbing a hill. The greater slopes result in more weight on the connecting edges. The end vertices of the edge contributes in the computation of the slope weight factor (say Sw). 4.2 Material Weight Factor The material present on ground as soft soil, sand, rocks, ice, snow, water, submerge etc. also plays an important role in deciding the mobility of a vehicle. Different kinds of vehicle are designed for different ground conditions. A particular vehicle can move on snow bound area easily but may not be good for movement in rocky ground. A set of rules can be devised for computation of material weight factor for different vehicle type movement on different material types. The material present on the end vertices of the edge participate in the computation of the

Saturn Software Mills Visit for more technical article http://www.saturnsoftmills.com/FreeDownloads.htm

material weight factor (say Mw). The total weight of the edge is computed as Tw = Sw + Mw 4.3 Weight Modifier Factor The roads and rivers may also be present in the operational area. It also affects the weight of the edge if the edge is passing through road or river. Roads are always preferred for vehicle movement on hilly terrain. If the edge is passing through the road (both end points lie on the road), this technique modifies the weight of the edge and assign it a very small weight. Tw 0 In contrast, if the edge is passing through the river, a very high weight is assigned to the edge to avoid the movement of the vehicle through the river. The weight of the edge is modified as ∞ Tw Using this technique, the weights of every edge is calculated in connected graph. This graph is mathematical representation of the mobility map. This graph can be converted in to a color image (mobility map) for human interpretation. Every ground cell is classified in to different categories depending on the weight values of eight edges originated from the cell. These categories may be represented by different colors in mobility map showing suitable regions for mobility of a particular vehicle type. References [1] Burton S. mobility”,ARMOR, 1982.

Boudinot, “Ground pp. 38-40, Jan-Feb

[2] A. Fournier & D.Y. Montuno, Triangulating simple polygons and equivalent problem, ACM Trans. Graphics, 1984, vol. 3, no. 2, 153-174. [3] Giorgio Franceschetti, Riccardo Lanari, “Synthetic Aperture Radar Processing”, CRC Press, Florida, 1999. [4] Y.C. Hsieh, D.M. McKeown & F.P. Perlant, Performance Evaluation of scene registration and stereo matching for cartographic feature extraction, IEEE Transaction on Pattern Analysis and Machine Intelligence, 1992, vol. 14, no. 2, 214-238. [5] John R. Jenson, “Introductory Digital Image Processing, A Remote Sensing Perspective”, Printice-Hall, Engleword Cliffs, New Jersey.