Next Article in Journal
Assessing the Spatial Accessibility of Urban Medical Facilities in Multi-Level and Multi-Period Scales Based on Web Mapping API and an Improved Potential Model
Previous Article in Journal
Thematic Content and Visualization Strategy for Map Design of City-Specific Culture Based on Local Chronicles: A Case Study of Dengfeng City, China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJGI
Volume 11
Issue 11
10.3390/ijgi11110544
ijgi-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Using Isovists in Measuring Surveillance and Expected Guardianship in Residential Neighborhood Property Crimes

by
Linda Nubani
1,* and
Jean Wineman
2
1
School of Planning, Design, and Construction, Michigan State University, East Lansing, MI 48824, USA
2
Taubman College of Architecture and Urban Planning, University of Michigan, Ann Arbor, MI 48109, USA
*
Author to whom correspondence should be addressed.
ISPRS Int. J. Geo-Inf. 2022, 11(11), 544; https://doi.org/10.3390/ijgi11110544
Submission received: 16 August 2022 / Revised: 3 October 2022 / Accepted: 15 October 2022 / Published: 31 October 2022
Browse Figures
Versions Notes

Abstract

:
Purpose: Assessing a level of surveillability, supervision, and expected guardianship in residential neighborhoods has been a topic of interest since the early work of Jacobs’ ‘eyes on the street’, and Newman’s ‘defensible space’. This paper reports on the use of isovists (two-dimensional polygons that represent the characteristics of the visual field) in understanding incidents of ‘breaking and entering’ in Ypsilanti, Michigan. Approach: Two measures relevant to environmental criminology were assessed: accessibility and surveillability. Findings: The findings indicate associations between incidents of crime and measures of visual accessibility. However, the level of homeownership was found to interact with the predictive models, suggesting the possible effect of ‘guardianship’. The geometrical shape of the isovist may also indicate where along a particular route, a crime is more likely to be committed. Originality: The results have the potential to assist law enforcement in identifying ‘hotspots’, and city planners in understanding the implications of urban design on crime.

1. Introduction

Crime prevention research has been a topic of interest since the development of the classical school of criminology in the 18th century. Although different in their approaches, early theories could be divided into two parts: those that focus on the actor and those that focus on the act. For example, theories from the classical school of criminology believed in deterrence as a way to prevent crime [1]. Offenders have the free will to commit their act and they go through a calculus method in weighing the risks of committing crimes. Other theories focused on the development of the criminal, and suggest factors that may influence this development, such as growing up in a socially disorganized neighborhood. Poverty and instability are all examples of a socially disorganized neighborhood [2,3]. Therefore, addressing the problems that lead to socially disorganized neighborhoods might be an effective means of crime prevention.
More recently, criminologists who focus on the act itself introduced new theories that linked crime with place. Examples include routine activity theory, rational choice theory, and crime prevention through environmental design (CPTED). These theories focus on the opportunities that lead to crime. Cohen and Felson [4], for example, suggest that suitable targets are often identified during the daily routines of a motivated offender. To locate suitable targets, Felson [5] argues that offenders will first look to ‘escape supervision’, next they search for locations that are not well maintained, and third they tend to look for places that lack supervision [6,7,8]. Examples of places that lack supervision include those that are occupied by renters, vacant lots, or abandoned structures [5]. Several studies also support the notion that residential stability represented by home ownership and length of stay has been shown to be crucial in reducing crime [2,9,10] and fear of crime [11,12,13], as it increases place attachment [14], collective efficacy [15,16], and sense of cohesion among neighbors [17].
Rational choice theory scholars argue that an offender’s rational choice is based on rewards and risks associated with their crime [18]. In a residential burglary, for example, being exposed to neighbors or passers-by will increase those risks, while the ability to access or leave the target property quickly will decrease those risks [19]. CPTED, which was introduced by criminologist C. Ray Jeffery [20], was later operationalized by Newman to include four environmental models [21] that are key to increasing those risks and enhancing levels of guardianship. These are: natural surveillance, access control, territoriality, and maintenance. Hollis, Felson and Welsh [22] (p. 76) define guardianship as “the presence of a human element which acts—whether intentionally or not—to deter the would-be offender from committing a crime against an available target”.
Non-criminologists that have focused their research on crime reduction through design include Jane Jacobs [23], and Oscar Newman [24]. Both Jacobs and Newman recognize the importance of natural surveillance. Jacobs’ theory encourages encounters with strangers to maintain constant ‘eyes on the street’, while Newman’s theory focuses on a separation between public and private areas so that residents can identify and report suspicious activities by strangers. To achieve each of these ends, Jacobs recommends mixing land uses, to encourage 24-h activity, and short urban neighborhood block lengths and multiple intersections, again, to increase movement and activity. Newman, on the other hand, favors a gradation in the hierarchy between private and public areas, where public areas have a clear separation from semi-private and private areas. Clear ownership is emphasized. Behavioral geographers, and more specifically chrono-urbanism advocates, have been examining ways to implement a concept known as the 15-min city, not only to reduce inequalities between different parts of cities, but to also be able to bring all services in a city within 15 min whether by bicycle or on foot, impact how people use space during the day and night, and increase their quality of life [25,26].
There is a considerable amount of research on measuring accessibility [27,28] and visibility [29,30] in relation to crime. Most of these studies focused on residential burglaries [31]. Some studies come with limitations [32]. First, some focused on the property, while others focused on the immediate environment of the property. Second, most studies focused on environmental features or sociodemographic characteristics, but not both. Third, there has not been a systematic way of measuring accessibility and visibility.
The research reported in this paper examines how spatial characteristics of the environment may contribute to incidents of criminal activity. The focus of the research was to objectively measure accessibility and visibility characteristics of street segments in residential neighborhoods in Ypsilanti, Michigan, in relation to incidents of breaking and entering, while controlling for sociodemographic variables. Specifically, the two key questions of the study are as follows: (1) what spatial characteristics appear to contribute to making target locations attractive to offenders? (2) can these spatial characteristics, such as visual access and visual control, be measured by isovist-based analysis?

2. Background Literature

2.1. Accessibility and Crime

Accessibility to potential targets as part of offenders’ routine activities has appeared in many studies by criminologists and planners. However, the definition of accessibility has varied across research groups [33,34]. In a major study by Weisburd, Groff and Yang [35], in addition to measuring sociodemographic variables (social disorganization theory), they also measured the accessibility of over 24,000 street segments in Seattle and compared them to over 1.6 million crime records (data from 1989 to 2004). In their study, they measured accessibility through the number of bus transportation links on a street segment and whether the street segment is considered to be an arterial road. They found that for each bus stop, the odds for crime hot spots on street segments double, and for each arterial road, the odds for crime hot spots increase more than ten times.
Other researchers have used different algorithms to measure accessibility. One such technique is drawn from space syntax theory, and measures accessibility based on the concept of visibility. Although there are different approaches and computer-based analysis programs to measure visibility (axial lines analysis, visual graph analysis (VGA), and others), they all begin with the 2-dimensional map of a city or selected area (or building). The space is then represented by sub-units (such as the representation of each street of the city by the fewest and longest straight lines of sight known as axial lines). Accessibility is then assessed using three measures: global integration, local integration, and connectivity. Global integration measures the accessibility of an axial line from all other axial lines in the city. Local integration looks at accessibility from a neighborhood of streets at a specified distance from the street segment (for example, three street segments away). Finally, connectivity examines the immediate adjacencies of the street segment—how many other street segments are directly connected to this segment. This method allows a researcher to objectively assess the likelihood that a certain street segment provides easy access to other locations, as compared to other street segments. When considering criminal activity, ease of access may facilitate both access to a target and escape opportunities from the target.
Earlier work by the authors of this paper documented the use of the lines analysis technique in calculating the accessibility level of 2117 axial lines in Ypsilanti, Michigan [36,37]. For this analysis, the authors controlled for sociodemographic variables including percentage of homeowners, youth composition, education, and income, all at a block-group level, as well as controlling for the length of line. Research results indicate that these factors had different effects when looking at the different types of crimes separately (the research looked at breaking and entering, larceny, motor vehicle theft, and robbery). When breaking and entering was examined, home ownership moderated the effect of both local integration and connectivity on breaking and entering (see Figure 1 to view the nature of this interaction). What these findings suggest is that locally integrated/connected routes (higher numbers of connections to other street segments in the local area) are associated with higher crime rates; however, with higher levels of homeownership, home burglary levels are lower. It may be that accessibility represents increased ease of escape routes, or it may be an indication that offenders are more likely to be introduced to their target during their routine activities [36,37]. When larceny was examined, length of line moderated the effect of connectivity on incidents of larceny. In the same research, similar results were confirmed when the author focused on three neighborhoods from the same city. This technique allowed the author to control for the length of line segment and the number of potential targets along a line [37].
Hillier and Sahbaz [38] introduced a new way to examine crime using the street segment as the unit of analysis. The researchers aggregated the segments according to the number of residences, then a burglary rate (instead of count) was established for each group of segments by dividing the number of burglaries by the number of residences for that group. The total number of groups (also referred to as risk bands) was 47. Their findings showed that burglary rates were high in integrated street segments with fewer residences and low in integrated street segments with more residences (indicating the importance of guardianship). Nes and Lopez [39] used the same techniques in two Dutch cities. A total of eight bands were constructed. In their study, space syntax measures of accessibility showed mixed results. Residential burglaries occurred on streets that are less integrated (quieter with less social control), while theft from vehicles occurred on streets that are more integrated (busier with quicker access to escape routes).
Summers and Johnson [40] also looked at the relationship between property crime and space syntax measures of accessibility to assess whether these measures are associated with the distribution of different types of violent crimes. They analyzed a total of 447 offenses within a 5-year period in London. Street segments (defined in their research as segments between intersections) were treated as the unit of analysis, and zero-inflated negative binomial regression was used due to the lack of crime associated with 95% of the street segments. Their findings showed that street segments that are high in integration, and segments close to those that are high in integration, were at higher risk of crime, indicating that the effect of through-movement extended to nearby segments. This research also found that if a segment was highly connected locally the risk of crime was lower.
Ward et al. [34] demonstrate inconsistencies among the findings of the studies that look at the relationship between accessibility and crime (specifically for burglary rates). Following traditional theories in criminology, Ward et al. [34] call for the importance of studying interactions between accessibility and socio-demographic variables, as there is some evidence that suggests that accessibility may have a different impact in different social contexts. For example, Matijosaitiene [41] found mixed results when comparing the spatial distribution of robberies and theft from motor vehicles in New Haven, CT, and Kaunas, Lithuania. Robberies and theft occurred on highly connected routes in New Haven, while in Kaunas similar crimes occurred on routes that were not necessarily highly integrated themselves, although they were connected to highly integrated routes. While the author attributed the difference in physical qualities in both cities to CPTED’s principles, it is unclear if robberies in both cities have similar attributes from a rational choice theory perspective. Numerous other studies found a positive association between different types of crime and integration [42,43]; however, none of these studies controlled for sociodemographic variables.
To this end, Ward et al. [34] examined burglary rates in 401 neighborhoods in Jacksonville, Florida. An accessibility or connectivity score was added to each neighborhood based on the ratio of the total number of street segments within that neighborhood to the total number of intersections. The researchers also created an index for several sociodemographic variables such as residential instability (percent homeownership) and concentrated disadvantage (percent black, percent unemployment, and percent below poverty line). Residential instability did not show an outcome effect in their model. Connectivity, however, was found to be related to increased burglary rates in low concentrated disadvantage neighborhoods, although it had an opposite effect in high concentrated disadvantage neighborhoods [31].
Kim and Choi [44] applied two types of statistical analysis (Poisson regression and negative binomial regression) to examine the relationship between space syntax measures and crime in Busan, China. In the model, they included CCTV ratio, female ratio, ratio of population over 65, the concentration of residents, and commercial units. Results showed that globally and locally integrated routes with a low number of connections were associated with crime. Not surprisingly, the presence of CCTV was associated with lower levels of crime. Further results showed that females and the population over 65 were more likely to become victims of crimes. The authors recommended that using negative binomial regression was more appropriate for this type of study. Research by di Bella et al. [45] followed a similar approach to analyze crimes across 83 streets in Genoa, Italy. They also found that integration was shown to be a statistically significant contributor to predatory crime, arson, and property damage. However, in their study, while the number of residents was found to be associated with higher crime levels, the number of shops and the length of the street were not significant.
In summary, there have been documented inconsistencies across the findings of space syntax studies and crime [34], most likely due to not controlling for length [35,40], not controlling for sociodemographic variables, and not examining different types of crimes separately [44]. Studies that focused on property crime used street segments as their unit of analysis, and integrated sociodemographic variables in their study generally found that property crimes occurred on highly connected or locally integrated street segments within less dense concentrated disadvantaged neighborhoods or neighborhoods with a lower percentage of homeownership [34,35,36,37].

2.2. Visibility and Crime

Benedikt (1979) introduced the concept of the ‘isovist’. This is a two-dimensional representation of the area that can be visually surveilled 360 degrees surrounding a point location. Davis and Benedikt [46] introduced a means to calculate and describe visual fields or ‘isovists’ mathematically. They presented a computational model for space representation where “a point x in a space P, the isovist at x, Vx, is the subset of P visible from x.” [47] (p. 49). Furthermore, they provided a detailed description of the algorithms used in calculating the different characteristics of isovists, such as area and perimeter. In the same paper, the authors recommended that since it has become possible to mathematically calculate isovists, this concept might be useful to the psychology and architectural research communities. More specifically, the authors called our attention to the possible link between isovists (visibility) and crime incidence. Criminals want to be safe from sudden detection by others and want to be inconspicuous.
An early work that described the relationship of visibility to crime appeared in the dissertation work of Camp [48] on bank robberies. Camp argued that most of the attention at the time was focused on the activity and capture of the criminal, rather than on the vulnerability of the target. Camp reported his findings of personally interviewing over 300 bank robbers [48] from a random sample of bank robbers confined in federal prisons. The robberies represented different banks from different cities and states. Camp augmented his notion on visual exposure and visual control in banks by comparing banks that were robbed to banks that were not robbed. Camp [48] drew the following conclusions: first, both the location and the physical characteristics of the bank were among the top reasons for robbers to select the bank; second, during the robbery, robbers attempted to blend with the crowd undisguised, and were concerned with being visible by the employees of the bank.
In 1985, Archea [49] proposed a theory of visual access and exposure, which suggests that as people move through space their visual access and exposure change. An area with high visual access provides high levels of awareness of the surroundings, and an area with high visual exposure give others the opportunity to monitor a person’s activity. Archea reviewed and analyzed the findings of previous studies on bank robberies and concluded that criminals selected target bank locations based on their visibility characteristics: high access (the ability of the perpetrator to continually monitor the surrounding space) and low exposure (locations where others might monitor the perpetrator).
Earlier use of isovist analysis in studies of crime appeared in a master’s thesis by Heard [50]. Heard examined several types of opportunity crimes (theft, larceny, as well as breaking and entering) in seven homogeneous locations (e.g., lot size, year built, wooden fences around the backyard, income) in Austin, Texas [50]. Heard generated 800 to 1000 isovist points within each block and registered the mean for different isovist properties. Results showed positive correlations between isovist properties (area, variance and dispersion) and incidents of crime, suggesting that crimes occur more often in/near spaces that lack visual control such as large parking lots and streets without direct residential surveillance.
Pereira Bezerra de Melo Junior and Canuto da Silva [51] used a similar approach to examine the isovist properties of locations that experienced thefts and robberies in a neighborhood in Recife, Brasil. Results of the correlational analysis showed that crimes occurred in areas with low occlusivity values, lower isovist area, lower isovist perimeter, and low visual integration. The authors suggest that criminals prefer places that allow them to have greater control of the visual field in an environment.
More recently, Aziz [52] constructed a visibility graph analysis for a 38,200-square meter park. All above-eye-level vegetation and other architectural elements were programmed to represent obstacles to sightlines. Findings showed that visually integrated and visually connected spaces recorded the lowest level of the types of crime (littering). Criminal activities like harassment, vandalism and littering occurred in medium visually integrated and connected areas. More serious crimes took place in areas with the least integration and connectivity.
Other studies that examined isovist properties focused on fear of crime. For example, Lee and Ha [53] used isovist-based analysis in exploring the relationship between visibility and fear of crime in six schools in the Songpa district of Seoul. A total of 1114 students were reported to be victims of crime within this school district. Based on data from 418 survey questionnaires, the authors identified a total of 858 locations where fear of crime was perceived to be high. Their findings indicated that areas with very high visual connectivity and areas with very low visual connectivity were strongly associated with higher fear of crime. Areas with very high visual connectivity were most often represented in the playgrounds of the schools, which could be an attraction for people to gather (and perhaps pose greater fear of crime), while areas with very low visual connectivity may lack the required guardianship to prevent crime from happening.
Kellom and Nubani [54] also performed a Visibility Graph Analysis on two university buildings. They measured faculty’s preparedness to execute the federally endorsed ‘Run, Hide, and Fight’ active shooter prevention training in relationship to their perception of wayfinding, layout complexity, and visual exposure. In addition to surveying faculty members with teaching assignments in these two buildings, six different law enforcement agencies participated in the survey. The findings suggest that isovist local and global measures were significantly associated with faculty’s perception of ease of wayfinding, and their assessment of whether the two building layouts were effective in their preparedness against active shooters. Law enforcement agencies emphasized how important it is for a building layout to have a hierarchy of visual access for officers to be able to clear the building room by room. These findings also echo Newman’s ideas of presenting a hierarchy of spaces from public to private [24].
Most of the past research that applied isovist-based analyses has centered on explaining the notion of wayfinding and people’s understanding of building layouts or street layouts spatially. From a criminology perspective, research explaining a criminal’s offending patterns using isovist-based analysis has great potential. This paper reports on the use of isovist-based analysis in exploring incidents of home burglaries (breaking and entering).

3. Methods

This exploratory research expands on the findings of an unpublished part of a doctoral dissertation that applied space syntax techniques to examine crime in Ypsilanti, Michigan [37]. As referenced in the background section, earlier findings suggest that locally integrated/connected routes (higher numbers of connections to other street segments in the local area) are associated with higher home burglary rates; however, with higher levels of homeownership, home burglary levels are lower. Other sociodemographic variables like youth concentration, median income levels, and racial composition were not predictive in this model. In this paper, results are presented on the relationship between isovist-based measures and home burglaries (breaking and entering). The study site was Ypsilanti, a city located within the Metropolitan Detroit area, with a population of approximately 22,362 at the time of the study. There were 338 breaking and entering incidents reported in the year 2003.
This study focuses on the spatial characteristics (visual access and visual control) of street segments, as measured by isovist-based analysis, and how these spatial characteristics may contribute to making target locations attractive to offenders. For this study, three neighborhoods were selected from Ypsilanti that were composed of single-family houses with similar architectural features. For example, the type and distribution of lighting across the three neighborhoods were constant. Road conditions and landscaping were similar. Access to amenities like parks, bus stops, and grocery stores was also similar. However, the overall spatial layout of each neighborhood differed. The first neighborhood is characterized by a semi-curvilinear layout. Approximately 75% of its residents are homeowners, with a medium income of USD 56,029. Only 8% of its residents are African Americans. The second neighborhood also has a semi-curvilinear layout. Approximately half of its residents are homeowners, with a median income between USD 19,000 and USD 29,000. Additionally, the majority of the residents are African Americans. The third neighborhood has a grid layout. Over 70% of the residents are homeowners, with a median income between USD 40,000 and USD 47,000. African Americans make up less than 9% of its residents. It is worth noting, however, that some of these neighborhoods are composed of two to three block groups. These neighborhoods were selected because they were the only three residential neighborhoods with single-family housing (to control for typology), and they were all within easy access to the main road.
Aerial photographs were used to trace the outlines of all structures using AutoCAD—a drafting software typically used by architectural and design-related disciplines. Moreover, the three neighborhoods were photographed and video documented to be able to identify above-eye-level physical elements such as trees, overgrown shrubs, and fences. All above-eye-level obstacles were traced using the same software. The outline of the three neighborhoods was later exported into two different software programs that perform different types of analysis: path isovist analysis and visibility graph analysis. Details of both procedures and programs are outlined below.

3.1. Path Isovist Analysis

As explained earlier, an isovist is the visible area surrounding a vantage point. In this study, a pathway was manually drawn for each accessible route within the three neighborhoods using Syntax 2D, a program developed by Turner et al. [55]. The path isovist is the analysis of the visual properties of a group of isovists set at a predetermined interval along a pathway. There was a total of 11 paths in the first neighborhood, 13 paths in the second neighborhood, and 16 paths in the third neighborhood. The program generated an isovist point along each pathway at approximately every 13 feet, a resolution that captured the average walking speed of approximately 13 feet every 3 s. All 40 pathways generated a total of 1048 isovist point locations. Figure 2 shows a sample of these pathways. This analysis was able to capture local measures from each isovist location such as area, perimeter, occlusivity and maximum length of the isovist radial. The last two measures are important in measuring visual access and exposure. Occlusivity expresses the ratio between occluded edges of vision and the amount of visible area from a vantage location. Higher occlusivity indicates that higher portions of the visual field are blocked (from view) by a visual barrier. Maximum isovist radial represents the longest line of sight within an isovist.

3.2. Visibility Graph Analysis

The authors also performed a visibility graph analysis on each of the three neighborhoods to understand the visual characteristics from a neighborhood level. To construct a visibility graph, a grid is overlaid covering the entire neighborhood area. Then isovists are generated from every grid intersection. The size of the grid (horizontal and vertical distances between the isovist locations) usually depends on the required resolution of analysis, the nature of the study (e.g., pedestrian movement vs. vehicular movement), and the required computing time. In this analysis, like the path analysis, isovist locations were also placed at an interval of 13 feet.
The visibility graph analysis was conducted using Depthmap 4.06, a program developed by Alasdair Turner at University College London [56]. Depthmap generated 27,890 grid points for the first neighborhood, 36,239 grid points for the second neighborhood, and 33,898 grid points for the third neighborhood. The program is designed to compute the visual relationship of these grid points to one another within the neighborhood. With such an advancement in technology, it has become possible to compute spatial properties of these isovists such as connectivity and integration.
In this database, the unit of analysis was the pathway (comparable to street segments). In addition to local isovist properties such as area and perimeter along each pathway, the visibility graph analysis allowed the researchers to attach neighborhood-level isovist properties such as connectivity and integration to each of the (grid-point) isovists that make up the path isovist. Sociodemographic variables were available at a block-group level and were amended to each pathway within the neighborhood (each neighborhood had an average of 2 to 4 block groups).

4. Results

The analysis was conducted using SAS9.4. The researchers examined the data for links between the different isovist measures and incidents of ‘breaking and entering’. A negative binomial regression analysis compared all 1084 isovist locations to incidents of breaking and entering in the three neighborhoods. PROC GENMOD was the procedure used for Generalized Estimating Equations (GEE) models. GENMOD offers the advantage of testing a sequence of models with one variable entered at a time until all the variables of interest are significant. Therefore, several models were tested for their adequacy using Criteria for Assessing Goodness of Fit. In the models, we tested several sociodemographic variables like homeownership, youth concentration, level of education, and median income. We also tested various visibility measures like isovist area and parameter, isovist compactness, as well as control and controllability. The model with the best fit showed that breaking and entering was associated positively with connectivity and the standard deviation of occlusivity, and negatively with the standard deviation of maximum isovist radial (see Table 1). It is worth noting that sociodemographic variables did not feature in the first model because percent ownership was found to be highly correlated with occlusivity (r = −0.33805, p = 0.03). When two variables are correlated (possible collinearity effect), it becomes difficult to assess the independent effect of each variable. Further statistical analysis also showed possible collinearity between income and home ownership (correlation coefficient 0.76, p < 0.0001), suggesting that homeowners have higher income than other groups. In a separate model, however, a negative association was found between percentages of homeowners and instances of breaking and entering (see Table 2).
These results suggest that breaking and entering will more likely occur along highly connected routes in block groups with lower percentages of homeowners. Additionally, target routes have higher occlusivity standard deviation and lower maximum isovist radial standard deviation than other routes. Since standard deviation measures how closely the values are to the mean, then a greater range of occlusivity indicates that as one moves along a pathway, there is a high variation in the amounts of ‘hidden’ views. This suggests patterns that potentially are not easily visually controlled from the street (eyes on the street) and might provide greater opportunities for concealment or escape. Maximum isovist radial deviation suggests that the maximum isovist radial, the longest view, changes as one moves along the path (the longest view is fairly unstable). The negative association to incidents of crime suggests that perpetrators may look for locations that give consistent (stable) longer views for surveillance. The researchers created box plots to demonstrate the difference between the standard deviation for both occlusivity and maximum radial length for the pathways that witnessed at least one crime against pathways that had no crime (see Figure 3). It was noted from analyzing the box plots that the box was comparatively short for pathways with crime, indicating that all the pathways within this group had similar standard deviation values that were high for occlusivity and low for isovist radial length. Research is recommended to further explore the impact of these variables on crime.
In addition to the use of negative binomial regression to analyze the data, the researchers augmented their analysis by examining the isovist characteristics and physical conditions around each crime incident location across all pathways. Using Google Earth 2D and 3D satellite imagery and videotapes of on-site visits to these neighborhoods, it was observed that occlusivity tends to vary largely based on the location of garages, and any gaps or other openings between walls/fences. Occlusivity tends to increase at locations where the garage of the house is located towards the back of the property (particularly for ones with narrow entrances). When the garage is located flush with the house, occlusivity tends to be lower. If some of the garages (particularly the ones with narrow entrances) were located behind the house, and others were located flush with the house on the same street, occlusivity would be varied and inconsistent. Figure 4 illustrates different scenarios. Moreover, occlusivity tends to be higher in areas around houses that offer more occluding edges that are ideal for concealment. Figure 5 is a good example of higher occlusivity. Gaps between a fence and a garage located behind a house will offer an occluding edge. On the other hand, Figure 6, is a good example of lower occlusivity created by a real-wall boundary behind the house, leaving no opportunities to hide behind wall surfaces.
Similarly, situations that tended to offer a greater range for maximum isovist radial were those with less consistency in long views. The negative association with breaking and entering suggests that perpetrators may look for areas where long views are fairly consistent, perhaps reflecting their need to visually control the area. Based on these results, houses could be safer from burglaries even if they sit on less locally connected streets, where there is not a lot of variability in obstructed views (occlusivity), and there are not consistent long views (maximum isovist radial).

5. Discussion

What have we learned from these analyses? The findings may offer a sense of comfort to neighborhoods with higher percentages of home ownership, even if they are located on locally well-connected streets. Homeowners are likely to show more territorial concerns and so suspicious activities are more likely to be reported. As earlier studies suggested [2,9,10,14], homeownership is associated with stability, familiarity, and security. Lower residential turnover is associated with increased social cohesion among neighbors. Homeowners who know more people tend to show more place attachment [14]. Streets that are more locally connected will likely generate more pedestrian traffic from within the neighborhood. These findings can be related to the effects of ‘expected levels of guardianship.’ If there are higher levels of homeownership (indicating a more stable population), under conditions of high and moderate levels of connectivity (supporting ‘more neighbors’, ‘defensible space’, and ‘eyes on the street’), breaking and entering is likely to be lower.
Physical characteristics around the houses are also important in facilitating the role of guarding one’s own neighborhood. To maximize levels of expected guardianship, it is important to recognize how visibility is bounded by the physical arrangement of objects within the space. This research found that incidents of crime are more likely to occur at locations along a street where there are significant changes in the isovist field—either a location that has a large increase or a large decrease in visibility as compared to the rest of the path. As seen in this research, gaps between fences and garages are examples of where these abrupt changes may occur. These locations could provide a concealment opportunity while allowing high visual control for the perpetrator and lower visual access for the guardians. Locations that have higher overall visibility, as compared to adjacent areas, may be identified as locations where the perpetrator has higher visibility (of potential observers) and easier routes of escape. These results are similar to the findings of Heard [50], who identifies open areas as potential target locations as well as locations where residential surveillance is limited.
These findings also echo Newman’s original ideas on controlling accessibility within the neighborhood. Neighborhoods should not act as a shortcut to get from point A to point B. If the street segments within these neighborhoods are globally integrated, it makes sense that from a burglar’s perspective that highly integrated routes in the city might provide introductions to potential targets. Within the local neighborhood, although homes on highly connected routes (local measure) may be more likely identified as targets simply because they tend to generate a lot of pedestrian movement and vehicular traffic, in support of Jacobs’ ideas, these connected routes also generate more ‘eyes on the street’. In neighborhoods along highly connected routes, with higher percentages of home ownership, occupants are more likely to experience lower levels of crime.
The exploratory analysis in this research brings together spatial, visual and sociodemographic factors in a single analytic framework. This type of analysis indicates that there are certain conditions that increase the chances for a breaking and entering event to occur at a certain place within a residential neighborhood. The analysis of breaking and entering showed that a crime will more likely occur on locally well-connected routes where obstructed views (occlusivity) are not consistent, long views (maximum isovist radial) are consistent, and there are low percentages of home ownership. Although the researchers integrated other important sociodemographic factors (that have appeared in research by social disorganization theorists) the sample size was too limited to demonstrate the role of these variables.

6. Conclusions

This exploratory study indicates that criminals are influenced by spatial considerations, and that among their concerns when selecting a target are environmental characteristics that affect accessibility, surveillability, and expected levels of guardianship. Accessibility is expressed by the connectivity of neighborhood routes, surveillability is measured by isovist properties, and guardianship is suggested by percentages of homeowners at a block-group level.
However, there is a need for further study as results of this analysis illuminated two potentially important issues in the study of crime: first, visibility attributes of streets are likely to be important for criminals’ decision making when they are in the process of selecting a target location; second, accessibility, visibility measures, and the sociodemographics of residential environments often interact, and should be studied together in the analysis of crime.
It was also shown in this study that both visual access and exposure could be measured objectively using space syntax measures such as isovist-based analysis. Some prior literature has pointed to the importance of access and exposure in the study of crime; however, these measures have not previously been easily operationalized. The difficulty in analyzing the extent of the influence of visual measures on crime is that there is often an overlap between the environments preferred by perpetrators and those preferred by occupants and other users. Visual access may give the passer-by a feeling of safety, while at the same time may provide the perpetrator a means to surprise the victim and an escape route. Further research on these potentially significant measures is needed. The analysis of visual measures in this research, however, is an initial step in applying objective measures to the concepts of access and exposure.

7. Implications for Future Research

Space syntax measures of accessibility and visibility appear to add a promising new tool for examining the effects of spatial layout characteristics and incidents of breaking and entering. This study makes three empirical contributions to crime-place studies: First, it supports major criminology theories such as routine activity theory, behavioral geography theory and social disorganization theory. It also suggests that spatial configuration has important implications with regards to home burglaries. However, the extent to which home burglaries are linked to these opportunities is also influenced by levels of homeownership. Therefore, it becomes important to analyze crime in the context of a robust model that considers sociodemographic, spatial and visual characteristics of street networks in a single framework. Thirdly, due to advancement in software, the research suggests that it is now possible to compute aspects of spatial configuration and visual characteristics that are relevant to understanding offenders’ preferences for target selection.

Author Contributions

Conceptualization, Linda Nubani and Jean Wineman; methodology, Linda Nubani and Jean Wineman; software, Linda Nubani; validation, Linda Nubani and Jean Wineman; formal analysis, Linda Nubani and Jean Wineman; investigation, Linda Nubani and Jean Wineman; resources, Linda Nubani and Jean Wineman; data curation, Linda Nubani and Jean Wineman; writing—original draft preparation, Linda Nubani; writing—review and editing, Linda Nubani and Jean Wineman; visualization, Linda Nubani; supervision, Jean Wineman; project administration, Linda Nubani All authors have read and agreed to the published version of the manuscript.”

Funding

This research received no external funding.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Freilich, J.D. Beccaria and situational crime prevention. Crim. Justice Rev. 2015, 40, 131–150. [Google Scholar] [CrossRef]
  2. Shaw, C.; McKay, H. Juvenile Delinquency and Urban Areas; Chicago Press, University of Chicago: Chicago, IL, USA, 1942. [Google Scholar]
  3. Kubrin, C.E.; Branic, N.; Hipp, J.R. (Re)conceptualizing Neighborhood Ecology in Social Disorganization Theory: From a Variable-Centered Approach to a Neighborhood-Centered Approach. Crime Delinq. 2021, 68, 2008–2032. [Google Scholar] [CrossRef]
  4. Cohen, L.E.; Felson, M. Social change and crime rate trends: A routine activity Approach. Am. Sociol. Rev. 1979, 44, 588–608. [Google Scholar] [CrossRef]
  5. Felson, M. Crime and Nature; Sage Publications: Thousand Oaks, CA, USA, 2006. [Google Scholar]
  6. Felson, M.; de Melo, S.N.; Boivin, R. Risk of outdoor rape and proximity to bus stops, bars, and residences. Violence Vict. 2021, 36, 723–738. [Google Scholar] [CrossRef] [PubMed]
  7. Moir, E.; Hart, T.C.; Reynald, D.M.; Stewart, A. Typologies of suburban guardians: Understanding the role of responsibility, opportunities, and routine activities in facilitating surveillance. Crime Prev. Community Saf. 2019, 21, 1–21. [Google Scholar] [CrossRef] [Green Version]
  8. Robinson, C.D. Routine Activity Theory and 13 Reasons Why: 13 Reasons Why: Routine Activity Theory. In Theories of Crime through Popular Culture; Daly, S.E., Ed.; Springer International Publishing: Cham, Switzerland, 2021; pp. 107–122. [Google Scholar]
  9. Lee, H.D.; Boateng, F.D.; Kim, D.; Binning, C. Residential Stability and Trust in the Police: An Understudied Area of Police Attitudinal Research. Am. J. Crim. Justice 2020, 45, 88–101. [Google Scholar] [CrossRef]
  10. Sozer, M.A.; Merlo, A.V. The Impact of Community Policing on Crime Rates: Does the Effect of Community Policing Differ in Large and Small Law Enforcement Agencies? Police Pract. Res. 2013, 14, 506–521. [Google Scholar] [CrossRef]
  11. Hardyns, W.; Snaphaan, T.; Pauwels, L.J.R.; Vyncke, V.; Willems, S. A Multilevel Analysis of Collective Efficacy, Neighborhood Disorder, and Individual Social Capital on Avoidance Behavior. Crime Delinq. 2019, 65, 994–1021. [Google Scholar] [CrossRef]
  12. 12 Lee, H.D.; Reyns, B.W.; Kim, D.; Maher, C. Fear of Crime OutWest: Determinants of Fear of Property and Violent Crime in Five States. Int. J. Offender Ther. Comp. Criminol. 2020, 64, 1299–1316. [Google Scholar] [CrossRef]
  13. Mrozla, T.; Dahle, T.O.; Hyunh, C.; Robinson, C.; Archibold, C.A.; Marcel, A. Fear of Crime in an Oil Boomtown in Western North Dakota. J. Crime Justice 2018, 41, 364–381. [Google Scholar] [CrossRef]
  14. Brown, G.; Brown, B.; Perkins, D.D. Place attachment in a revitalizing neighborhood: Individual and block levels of analysis. J. Environ. Psychol. 2003, 23, 259–271. [Google Scholar] [CrossRef]
  15. Swatt, M.L.; Varano, S.P.; Uchida, C.D.; Solomon, S.E. Fear of Crime, Incivilities, and Collective Efficacy in four Miami Neighborhoods. J. Crim. Justice 2013, 41, 1–11. [Google Scholar] [CrossRef] [Green Version]
  16. Yuan, Y.; McNeeley, S. Social Ties, Collective Efficacy, and Crime-Specific Fear in Seattle Neighborhoods. Vict. Offenders 2015, 12, 90–112. [Google Scholar] [CrossRef]
  17. Gau, J.; Corsaro, N.; Stewart, E.; Brunson, R. Examining Macro-Level Impacts on Procedural Justice and Police Legitimacy. J. Crim. Justice 2012, 40, 333–343. [Google Scholar] [CrossRef]
  18. Cornish, D.B.; Clarke, R.V. The Reasoning Criminal: Rational Choice Perspectives on Offending; Springer: Secaucus, NJ, USA, 1986. [Google Scholar]
  19. Lee, K.H.; Kim, J.W. An Experimental Study on the Environmental Cues which affect Target Searching and Selection Process of Residential Burglars. J. Archit. Inst. Korea 2000, 16, 83–90. [Google Scholar]
  20. Jeffery, C.R. Crime Prevention through Environmental Design; Sage Publications: Beverly Hills, CA, USA, 1971. [Google Scholar]
  21. Jeffery, C.R. CPTED, Past, Present and Future. Int. Crime Prev. through Environ. Des. Newsl. 1999, 3, 1–8. [Google Scholar]
  22. Hollis, M.E.; Felson, M.; Welsh, B.C. The capable guardian in routine activities theory: A theoretical and conceptual reappraisal. Crime Prev. Community Saf. 2013, 15, 65–79. [Google Scholar] [CrossRef]
  23. Jacobs, J. The Death and Life of Great American Cities; Random House: New York, NY, USA, 1961. [Google Scholar]
  24. Newman, O. Defensible Space; Macmillan: New York, NY, USA, 1972. [Google Scholar]
  25. Marino, M.; Tomaz, E.; Henriques, C.; Chavoshi, S.H. The 15-minute city concept and new working spaces: A planning perspective from Oslo and Lisbon. Eur. Plan. Stud. 2022, 1–23. [Google Scholar] [CrossRef]
  26. Mocak, P.; Matlovicova, K.; Matlovic, R.; Penzes, J.; Pachura, P.; Mishra, P.K.; Kostilnikova, K.; Demkova, M. 15-Minute City Concept as a Sustainable Urban Development Alternative: A Brief Outline of Conceptual Frameworks and Slovak Cities as a Case. Folia Geogr. 2022, 64, 69–89. [Google Scholar]
  27. Armitage, R. Burglars’ take on crime prevention through environmental design (CPTED): Reconsidering the relevance from an offender perspective. Secur. J. 2018, 31, 285–304. [Google Scholar] [CrossRef] [Green Version]
  28. Groff, E.R.; LaVigne, N.G. Mapping an Opportunity Surface of Residential Burglary. J. Res. Crime Delinq. 2001, 38, 257–278. [Google Scholar] [CrossRef]
  29. Armitage, R. Domestic burglary: Burglar responses to target attractiveness. In Reducing Burglary; Tseloni, A., Thompson, R., Tilley, N., Eds.; Springer: Cham, Switzerland, 2018; pp. 45–75. [Google Scholar]
  30. Armitage, R.; Joyce, C. Why My House?’: Exploring the Influence of Residential Housing Design on Burglar Decision-Making. In Re-building Crime Prevention through Environmental Design (CPTED); Armitage, R., Ekblom, P., Eds.; Taylor and Francis: Oxfordshire, UK, 2017. [Google Scholar]
  31. Nobles, M.; Ward, J.; Tillyer, R. The impact of neighborhood context on spatiotemporal patterns of burglary. J. Res. Crime Delinq. 2016, 53, 711–740. [Google Scholar] [CrossRef]
  32. Montoya, L.; Junger, M.; Ongena, Y. The Relation Between Residential Property and Its Surroundings and Day- and Night-Time Residential Burglary. Environ. Behav. 2016, 48, 515–549. [Google Scholar]
  33. Coppola, F.; Grimaldi, M.; Fasolino, I. A Configurational Approach for Measuring the Accessibility of Place as an Analysis Tool for Crime Risk Vulnerability. Innov. Urban Reg. Plan. 2021, 146, 501–509. [Google Scholar]
  34. Ward, J.T.; Nobles, M.R.; Youstin, T.J.; Cook, C.L. Placing the neighborhood Accessibility–Burglary link in social-structural context. Crime Delinq. 2014, 60, 739–763. [Google Scholar] [CrossRef]
  35. Weisburd, D.; Groff, E.R.; Yang, S. Understanding and controlling hot spots of crime: The importance of formal and informal social controls. Prev. Sci. 2014, 15, 31–43. [Google Scholar] [CrossRef]
  36. Nubani, L.; Wineman, J. The Role of Space Syntax in Identifying the Relationship between Space and Crime. In Proceedings of the Fifth International Space Syntax Symposium, Delft, The Netherlands, 13–17 June 2005; van Nes, A., Ed.; Techne Press: Delft, The Netherlands, 2005; pp. 413–422. [Google Scholar]
  37. Nubani, L. Targets for Crime: Measuring the Spatial and Visual Attributes of Crime Locations Using Space Syntax. Ph.D. Thesis, University of Michigan, Ann Arbor, MI, USA, 2006. [Google Scholar]
  38. Hillier, B.; Sahbaz, O. High resolution analysis of crime patterns in urban street networks: An initial statistical sketch from an ongoing study of a London borough. In Proceedings of the Fifth International Space Syntax Symposium, Delft, The Netherlands, 13–17 June 2005; van Nes, A., Ed.; Techne Press: Delft, The Netherlands, 2005; pp. 451–478. [Google Scholar]
  39. Van Nes, A.; López, M. Macro and micro scale spatial variables and the distribution of residential burglaries and theft from cars: An investigation of space and crime in the Dutch cities of Alkmaar and Gouda. J. Space Syntax 2010, 1, 314. [Google Scholar]
  40. Summers, L.; Johnson, S.D. Does the configuration of the street network influence where outdoor serious violence takes place? using space syntax to test crime pattern theory. J. Quant. Criminol. 2017, 33, 397–420. [Google Scholar] [CrossRef] [Green Version]
  41. Matijosaitiene, I. Combination of CPTED and space syntax for the analysis of crime. Safer Communities 2016, 15, 49–62. [Google Scholar] [CrossRef]
  42. De Filippi, F.; Cocina, G.G.; Martinuzzi, C. Integrating Different Data Sources to Address Urban Security in Informal Areas. The Case Study of Kibera, Nairobi. Sustainability 2020, 12, 2437. [Google Scholar] [CrossRef] [Green Version]
  43. Sahajramani, D.; Purkayastha, S.; Ranjit, J.; Vyas, A. Determination of correlation between street accessibility and crimes using space syntax network graph analysis. In Proceedings of the 39th Asian Conference on Remote Sensing, Kuala Lumpur, Malaysia, 15–19 October 2018; Asian Association on Remote Sensing (AARS): Kuala Lumpur, Malaysia, 2018. [Google Scholar]
  44. Kim, H.J.; Choi, Y. A study on the influence of the space syntax and the urban characteristics on the incidence of crime using negative binomial regression. KSCE J. Civ. Environ. Eng. Res. 2016, 36, 333–340. [Google Scholar]
  45. di Bella, E.; Corsi, M.; Leporatti, L.; Persico, L. The spatial configuration of urban crime environments and statistical modeling. Environ. Plan. B Urban Anal. City Sci. 2017, 44, 647–667. [Google Scholar] [CrossRef]
  46. Benedikt, M.L. To take hold of space: Isovists and isovist fields. Environ. Plan. B 1979, 6, 47–65. [Google Scholar] [CrossRef]
  47. Davis, L.S.; Benedikt, M.L. Computational models of space: Isovists and isovist Fields. Comput. Graph. Image Process. 1979, 11, 49–72. [Google Scholar] [CrossRef]
  48. Camp, G.M. Nothing to Lose: A Study of Bank Robbery in America. Ph.D. Thesis, Yale University, New Haven, CT, USA, 1968. [Google Scholar]
  49. Archea, J.C. The use of architectural props in the conduct of criminal acts. J. Archit. Plan. Res. 1985, 2, 245–259. [Google Scholar]
  50. Heard, L. Space, Place and Opportunistic Crime: A Study Using Isovists and Space Syntax. Master’s Thesis, University of Texas at Austin, Austin, TX, USA, 2004. [Google Scholar]
  51. Pereira Bezerra de Melo Junior, S.; Canuto da Silva, R. Applicability of 2D and 3D isovists and visibility graph analysis for evaluating urban vulnerability to crime: The case of Boa Viagem, in Recife. In Proceedings of the 22nd Conference of the Iberoamerican Society of Digital Graphics, São Carlos, Brazil, 7–9 November 2018; pp. 308–315. [Google Scholar]
  52. Aziz, N.A.A.E. Space Syntax as a Tool to Measure Safety in Small Urban Parks—A Case Study of Rod El Farag Park in Cairo, Egypt. Landsc. Archit. Front. 2020, 8, 42–59. [Google Scholar] [CrossRef]
  53. Lee, S.; Ha, M. The duality of visibility: Does visibility increase or decrease the fear of crime in schools′ exterior environments? J. Asian Archit. Build. Eng. 2015, 14, 145–152. [Google Scholar] [CrossRef] [Green Version]
  54. Kellom, K.; Nubani, L. One Step Ahead of Active Shooters: Are Our University Buildings Ready? Buildings 2018, 8, 173. [Google Scholar] [CrossRef]
  55. Turner, J.; Wineman, J.; Psarra, S.; Jung, S.K.; Senske, N. Syntax 2D 1.3.0.7; University of Michigan: Ann Arbor, MI, USA, 2006. [Google Scholar]
  56. Turner, A. Depthmap 4.06; University College London: London, UK, 2004. [Google Scholar]
Ijgi 11 00544 g001 550
Figure 1. Plot of the interaction of connectivity and homeownership on breaking and entering. Note: When connectivity is high and percentage of homeownership is high, crime rate is lower.
Figure 1. Plot of the interaction of connectivity and homeownership on breaking and entering. Note: When connectivity is high and percentage of homeownership is high, crime rate is lower.
Ijgi 11 00544 g001
Ijgi 11 00544 g002 550
Figure 2. A sample path isovist output in the first neighborhood. The shaded areas represent the total visibility or amount of visual information available when one takes that route.
Figure 2. A sample path isovist output in the first neighborhood. The shaded areas represent the total visibility or amount of visual information available when one takes that route.
Ijgi 11 00544 g002
Ijgi 11 00544 g003 550
Figure 3. Boxplots comparing the standard deviation for Isovist Max Radial and the standard deviation for Occlusivity for paths containing crime against paths with no crime in all three neighborhoods.
Figure 3. Boxplots comparing the standard deviation for Isovist Max Radial and the standard deviation for Occlusivity for paths containing crime against paths with no crime in all three neighborhoods.
Ijgi 11 00544 g003
Ijgi 11 00544 g004 550
Figure 4. These four scenarios illustrate the positioning of garages in relation to the house.
Figure 4. These four scenarios illustrate the positioning of garages in relation to the house.
Ijgi 11 00544 g004
Ijgi 11 00544 g005 550
Figure 5. The overall sketch illustrates how the fence at the back of the house is flush with the garage. This is a good example of how to keep occlusivity low.
Figure 5. The overall sketch illustrates how the fence at the back of the house is flush with the garage. This is a good example of how to keep occlusivity low.
Ijgi 11 00544 g005
Ijgi 11 00544 g006 550
Figure 6. The overall sketch illustrates how the fence at the back of the house creates an opportunity for reduced informal guardianship. This is a good example of having increased values in occlusivity.
Figure 6. The overall sketch illustrates how the fence at the back of the house creates an opportunity for reduced informal guardianship. This is a good example of having increased values in occlusivity.
Ijgi 11 00544 g006
Table
Table 1. Output of negative binomial regression representing the effects of connectivity and visual measures on breaking and entering.
Table 1. Output of negative binomial regression representing the effects of connectivity and visual measures on breaking and entering.
Analysis of GEE Parameter Estimates
Empirical Standard Error Estimates
ParameterEstimateStandard ErrorPr > |Z|
Intercept−0.44940.0108<0.0001
Connectivity0.29860.09720.0021
Std. Dev. Occlusivity0.27070.0423<0.0001
Std. Dev. Max Isovist Radial−0.45770.0265<0.0001
Table
Table 2. Output of negative binomial regression representing the effects of homeownership on breaking and entering.
Table 2. Output of negative binomial regression representing the effects of homeownership on breaking and entering.
Analysis of GEE Parameter Estimates
Empirical Standard Error Estimates
ParameterEstimateStandardPr > |Z|
Error
Intercept0.41470.12060.0006
Homeownership−1.14650.1363<0.0001
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Nubani, L.; Wineman, J. Using Isovists in Measuring Surveillance and Expected Guardianship in Residential Neighborhood Property Crimes. ISPRS Int. J. Geo-Inf. 2022, 11, 544. https://doi.org/10.3390/ijgi11110544

AMA Style

Nubani L, Wineman J. Using Isovists in Measuring Surveillance and Expected Guardianship in Residential Neighborhood Property Crimes. ISPRS International Journal of Geo-Information. 2022; 11(11):544. https://doi.org/10.3390/ijgi11110544

Chicago/Turabian Style

Nubani, Linda, and Jean Wineman. 2022. "Using Isovists in Measuring Surveillance and Expected Guardianship in Residential Neighborhood Property Crimes" ISPRS International Journal of Geo-Information 11, no. 11: 544. https://doi.org/10.3390/ijgi11110544

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop