**4. Proposing novel methodology and 3D metrics for sustainable urban form assessment over time**

#### **4.1 Airborne lidar and technical methods**

Light detection and ranging (lidar) measures ranges from an airborne scanner which scans at right angles to the flight direction, to the terrain surface using a pulsed laser and generates rich 3D point clouds about the terrain surface and objects on the surface. The round trip time of the laser pulses from aircraft to the ground allows the determination of the distance from the laser scanner to the terrain or objects. Together with known positions and attitude of the scanner derived from a global navigation satellite system (GNSS) receiver for the determination of instantaneous positions of the aircraft and an inertial measuring unit (IMU) for the determination of instantaneous velocities and orientations of the aircraft this also enables determination of accurate 3D coordinates of points representing the terrain or any visible objects on the terrain. After pre-processing, the data can be used for a wide range of applications such as urban growth monitoring, biomass estimation and object measurements. **Figure 2** illustrates the summary of the entire process of data acquisition and applications.

A digital surface model (DSM) is generated from airborne lidar data elevation values projected on a raster data format.

**Figure 3** represents a flowchart of our tested time series airborne lidar data processing and the method for application of the proposed 3D metrics. Ground, vegetation and building classes can be extracted from temporal airborne lidar data sets which are Digital Surface Models (DSM). The difference between the ground elevation and elevations of buildings and vegetation, provides the height of these objects above ground. Also, voxelization of the buildings and vegetation classes would be advantageous to overcome the problem of different point density and sampling of the time series airborne lidar data. Comparison of the buildings and vegetation changes over time from either approaches (i.e. the approach of

#### **Figure 2.**

*Generating big data for urban growth and construction change analysis, (a) airborne lidar; (b) biomass survey; (c) dimensions of buildings; (d) data collection and processes.*

**59**

large urban areas.

**Figure 3.**

in the building class.

added for this purpose.

**4.2 Building and vegetation metrics**

*4.1.1 Study extents*

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time…*

classification then detecting changes or the approach of change detection) can be used in a SUF study so as to see how the volumes of buildings change compared to

*The flowchart of the proposed procedure of airborne lidar data processing for the calculation of 3D metrics of sustainable urban form. Note: UFS: Urban form sustainability; DBM: Digital building model; DVM: Digital* 

One of the challenges regarding various change detection algorithms is that, especially for time series airborne lidar, the performances of these algorithms are not evaluated for large data sets covering vast urban areas. In addition, the results of these algorithms are not compared to identify the most appropriate one for determining three-dimensional changes of urban developments. Matikainen et al. [26] have also confirmed that there have been limited studies on change detection over

A lack of criteria on which to base comparisons of change detection methods for characterizing vertical urban development is also another problem. In this study, two criteria for the evaluation of change detection algorithms have been developed; namely, determination of the range of height changes; and differentiation between new or increased heights, as against demolished or decreased heights

Time series airborne lidar data for the University of New South Wales are considered as spatiotemporal 3D test data to quantify 3D changes of buildings compared with vegetation. **Figure 4a–c** illustrates these data sets, and areas of 3D changes can be seen by comparing them. Details of these data sets are summarized in **Table 1**.

As discussed before, there is a lack of metrics for assessment of sustainability of urban form as a 3D phenomenon over time. We propose the following metrics to be

the volume of vegetation, as one indicator of sustainability.

*vegetation model; bld: Building; veg: Vegetation.*

*DOI: http://dx.doi.org/10.5772/intechopen.89617*

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time… DOI: http://dx.doi.org/10.5772/intechopen.89617*

#### **Figure 3.**

*Sustainability in Urban Planning and Design*

**form assessment over time**

**4.1 Airborne lidar and technical methods**

entire process of data acquisition and applications.

values projected on a raster data format.

vegetation.

Considering current rapid urbanization and the above-mentioned lack of studies on sustainable 3D urban development, there is an urgent need to numerically characterize how 3D urban development affect the sustainability goals of caring for the environment, including maintaining (and ideally, enhancing)

**4. Proposing novel methodology and 3D metrics for sustainable urban** 

Light detection and ranging (lidar) measures ranges from an airborne scanner which scans at right angles to the flight direction, to the terrain surface using a pulsed laser and generates rich 3D point clouds about the terrain surface and objects on the surface. The round trip time of the laser pulses from aircraft to the ground allows the determination of the distance from the laser scanner to the terrain or objects. Together with known positions and attitude of the scanner derived from a global navigation satellite system (GNSS) receiver for the determination of instantaneous positions of the aircraft and an inertial measuring unit (IMU) for the determination of instantaneous velocities and orientations of the aircraft this also enables determination of accurate 3D coordinates of points representing the terrain or any visible objects on the terrain. After pre-processing, the data can be used for a wide range of applications such as urban growth monitoring, biomass estimation and object measurements. **Figure 2** illustrates the summary of the

A digital surface model (DSM) is generated from airborne lidar data elevation

**Figure 3** represents a flowchart of our tested time series airborne lidar data processing and the method for application of the proposed 3D metrics. Ground, vegetation and building classes can be extracted from temporal airborne lidar data sets which are Digital Surface Models (DSM). The difference between the ground elevation and elevations of buildings and vegetation, provides the height of these objects above ground. Also, voxelization of the buildings and vegetation classes would be advantageous to overcome the problem of different point density and sampling of the time series airborne lidar data. Comparison of the buildings and vegetation changes over time from either approaches (i.e. the approach of

*Generating big data for urban growth and construction change analysis, (a) airborne lidar; (b) biomass* 

*survey; (c) dimensions of buildings; (d) data collection and processes.*

**58**

**Figure 2.**

*The flowchart of the proposed procedure of airborne lidar data processing for the calculation of 3D metrics of sustainable urban form. Note: UFS: Urban form sustainability; DBM: Digital building model; DVM: Digital vegetation model; bld: Building; veg: Vegetation.*

classification then detecting changes or the approach of change detection) can be used in a SUF study so as to see how the volumes of buildings change compared to the volume of vegetation, as one indicator of sustainability.

One of the challenges regarding various change detection algorithms is that, especially for time series airborne lidar, the performances of these algorithms are not evaluated for large data sets covering vast urban areas. In addition, the results of these algorithms are not compared to identify the most appropriate one for determining three-dimensional changes of urban developments. Matikainen et al. [26] have also confirmed that there have been limited studies on change detection over large urban areas.

A lack of criteria on which to base comparisons of change detection methods for characterizing vertical urban development is also another problem. In this study, two criteria for the evaluation of change detection algorithms have been developed; namely, determination of the range of height changes; and differentiation between new or increased heights, as against demolished or decreased heights in the building class.

#### *4.1.1 Study extents*

Time series airborne lidar data for the University of New South Wales are considered as spatiotemporal 3D test data to quantify 3D changes of buildings compared with vegetation. **Figure 4a–c** illustrates these data sets, and areas of 3D changes can be seen by comparing them. Details of these data sets are summarized in **Table 1**.

#### **4.2 Building and vegetation metrics**

As discussed before, there is a lack of metrics for assessment of sustainability of urban form as a 3D phenomenon over time. We propose the following metrics to be added for this purpose.

#### **Figure 4.**

*Time series airborne lidar data over the selected area of Sydney, Australia and demonstrated areas of changes in yellow and red rectangles: (a) lidar data collected in 2005; (b) lidar data collected in 2008 and (c) lidar data collected in 2013.*


**61**

**5. Results**

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time…*

For comparison of the changes of new and demolished buildings, a volumetric

∑*i*=1 *<sup>l</sup> d*× *hid*

(1)

*Vc* (2)

*Vc* (3)

*RV C*= ∑*i*=1 *k* \_ *n*× *hin*

where *n* demonstrates the area of a pixel on detected new buildings and *hin* demonstrates the height of that pixel above ground, *d* represents the area of a pixel in demolished buildings and *hid* is its corresponding height value. This ratio can be

One of the limitations of DSM differencing method is the selection of the appropriate thresholds for the unchanged class. Lu et al. [27] proposed thresholds to be [*m* − ,*m* + ] in the histogram of the resultant difference image (i.e. *DSMd*), where *m* is the mean of the distribution and σ represents the standard deviation. Shirowzhan [28] tested different values between 2 and 4 for the thresholds of γ in

The next step for this 3D metric is a careful visual inspection to determine whether the areas of change belong to buildings because the result of image differencing includes both buildings and vegetation changes. If this ratio is more than 1, it means that the volume of new buildings is greater than the volume of demolished buildings. However, this metric cannot show whether the development is of vertical form. Therefore, complementary descriptors are necessary to recognize whether the urban form change is vertical growth. For example, if this ratio is greater than 1 and the new buildings are higher than the demolished ones, it can be concluded that an

Other volumetric descriptors for spatial and temporal 3D urban growth can be proposed which use improved concepts of 3D mass and space in urban planning. Examples of a 3D mass index and a 3D space index are defined in (2) and (3) as the ratio of the volume of buildings (or vacant space if applicable) to the total volume of an assumed cube, whose footprint is the ground surface area including all buildings, and its height equals the tallest building height in the study extent. For a study extent, after calculation of the volume of new and demolished buildings, an increase in the 3D mass index will represent a trend toward more compact 3D urban form. The assumption for the cube is that it only contains buildings.

*Vb Vc* = ∑*i*=1 *n* \_ *b*× *hb*

*Vs*

where *Vb* is the volume of buildings, *Vc* is the volume of the containing cube and *Vs* represents the volume of vacant space. In a pixel-based method, *b* is the area of a pixel representing built form and *hb* is the corresponding height of the pixel.

Temporal point clouds of the selected case study were analyzed based on the developed procedure (refer to **Figure 3**). The purpose of the analysis was vegetation

*Vc* = *Vc* <sup>−</sup>∑*i*=1

*n* \_\_\_\_\_\_\_\_\_\_\_ *<sup>b</sup>*× *hb*

3D *MassIndex* = \_

3D *SpaceIndex* = \_

descriptor of ratio of volume change (*RVC*) is proposed in (1):

calculated using the results of image differencing.

this study area and found that = 3 is an appropriate value.

infill vertical development is occurring in a study area.

*4.2.2 Mass and space index*

*DOI: http://dx.doi.org/10.5772/intechopen.89617*

*4.2.1 Ratio of volume change (RVC)*

#### **Table 1.**

*Detailed information of airborne lidar data sets used in this study.*

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time… DOI: http://dx.doi.org/10.5772/intechopen.89617*
