**3. The WSL monoplotting tool**

The implementation of the monoplotting principle in a practical tool has been restricted by the lack of basic data (i.e., detailed digital elevation models [26]) or the inadequacy of the available computing power [27]. As a result, the pioneering work of Makarovič [25, 28] has long remained isolated. Only in the last 20 years have new software and tools been developed based on the monoplotting principle [29–32]. None of these products, however, really meet the needs of potential end users in terms of operational flexibility and user-friendliness, thus, greatly inhibiting their broad use.

Recently, improvements in digital photography (e.g., high-resolution digital cameras, digitalization of historical pictures in high-resolution) and advancements in computing science have opened new possibilities for developing a specific monoplotting tool with an interface that makes it easy to handle not only by specialized researchers, but also non-expert users for operational purposes [21]. To that end, we developed the WSL monoplotting tool (available at present in the 2.0 version) [24], which has the following main features and characteristics (see also https://www.wsl.ch/monoplotting for detail):


Model: DTM) or may include vegetation, buildings, and other vertical objects (Digital Sur-

• **Control points (CPs)**, which are precisely and unambiguously identifiable locations (e.g., road and footpath intersections, rocky outcrops, and building corners) on both the image (pixel coordinates) and the landscape (real world coordinates, i.e., latitude, longitude, and altitude of the real-world coordinates). CPs should be at least four or more in number, preferably placed on the ground DTM, and possibly homogeneously distributed over the entire image. The corresponding real-world coordinates may be derived from georeferenced geographical information (e.g., orthophotos, maps, cadaster, DTM, and DSM) or can

Once calibrated in such a monoplotting system, single oblique pictures can also be interpreted three-dimensionally (3D). The main limitation of the monoplotting system is the fact that in mono-photogrammetry, only points on the terrain (DEM-surface) can be precisely located whereas stereo-photogrammetry enables the calculation of the position of any common pixel

The implementation of the monoplotting principle in a practical tool has been restricted by the lack of basic data (i.e., detailed digital elevation models [26]) or the inadequacy of the available computing power [27]. As a result, the pioneering work of Makarovič [25, 28] has long remained isolated. Only in the last 20 years have new software and tools been developed based on the monoplotting principle [29–32]. None of these products, however, really meet the needs of potential end users in terms of operational flexibility and user-friendliness, thus,

Recently, improvements in digital photography (e.g., high-resolution digital cameras, digitalization of historical pictures in high-resolution) and advancements in computing science have opened new possibilities for developing a specific monoplotting tool with an interface that makes it easy to handle not only by specialized researchers, but also non-expert users for operational purposes [21]. To that end, we developed the WSL monoplotting tool (available at present in the 2.0 version) [24], which has the following main features and characteristics (see

• A user-friendly, intuitive, and self-explanatory interface enabling the simultaneous visualization of one or more photographs of the target landscape as well as of related orthopho-

• A computer-assisted, semiautomatic, and interactive calibration process for the camera, including the reconstruction of all elements in the monoplotting system (e.g., snapshot

• An immediate estimate of the error for each control point used for the calibration of the

tos, maps, or other georeferenced representation of the terrain surface (**Figure 2**).

be directly surveyed in the field instrumentally (e.g., GPS).

face Model: DSM).

110 Natural Hazards - Risk Assessment and Vulnerability Reduction

in the stereo pair.

location).

oblique photograph.

**3. The WSL monoplotting tool**

greatly inhibiting their broad use.

also https://www.wsl.ch/monoplotting for detail):

• Export–import routines allowing data exchange (e.g., CSV or shapefiles) between and from conventional geographic information systems (GIS).

The heart of the tool is the iterative process whereby camera calibration is achieved. This is done in order to precisely estimate and simulate the three intrinsic and the six extrinsic camera parameters. The three intrinsic parameters are the two-pixel coordinates (xc, yc) of the principal point (i.e., the image center), and the principal distance (perpendicular distance from the image to the projection center). The six extrinsic parameters are the three real-word coordinates of the lens pinhole and the three Euler rotation angles α (pan/z-axis), β (tilt/yaxis), and γ (roll/x-axis) of the camera (see [24, 33] for more details). The calibration routine generates a sequence of collinearity equations commonly used in photogrammetry [34] that estimates and approximates the unknown camera parameters [35] in order to progressively minimize the error of the camera model when applied to the input data. Once camera calibration is achieved, the tool implements a model of all the extrinsic and intrinsic camera parameters, which simulates the original camera setup when the picture was taken. The tool also

**Figure 2.** The MPT\_2.0 applied to the Sommascona flood event (see Example 2). A major strength of the 2.0 release is the option of opening and processing more oblique images or maps of the same area to obtain a complete view of the event.

returns the theoretical 3D errors, which correspond to the deviation between the real-world coordinates of each control point and the corresponding coordinates as calculated from the calibration procedure of the monoplotting system.

It is important to note that shooting point reconstruction is implemented so as to optimize the overlap of the corresponding CPs in the image to be georeferenced and in the georeferenced map, respectively. Thus, the reconstructed theoretical shooting point may not necessarily correspond precisely to the real camera position, and the two may differ by a few centimeters to several meters.
