Remote Sensing Assignment On “Crime Mapping And Analysis”
Modern computer technologies and availability of easy and cheap spatial and temporal data has paved way for analysing and forecasting various patterns easily- crime being one of them. Satistical tools like those for hotspot analysis are of help in finding crime hotspots. The study at hand takes into consideration three such tools, namely Kernel Density Estimation, Inverse Distance Weighted and Getis-Ord Gi* methods and are applied on New Delhi, the capital of India, for studying crimes and their locations such that proper actions can be taken timely
Introduction GIS is the computer system designed for capturing, preserving, querying, exploring and displaying geospatial data, which shows both locational data and its attributes. Crime needs no specific definition. Its impact has been rising enormously over the past years. The Police is in its constant struggle to curb the incidences, in an effective manner; and to increase the effectiveness, it is necessary to analyse and identify the reasons behind and the areas where such notorious actions are taking place. In order to carry out this huge task, it is necessary to make use of modern technologies line GIS. Crime Mapping and Analysis is one such way, which makes use of the locations where the various crimes are taking place. As a result, solutions can be provided, like improved patrolling. There is an urgent need to analyze the increasing number of crimes as approximately 17 lakhs Indian Penal Code (IPC) crime, and 38 lakhs local and Special Law crimes per year.
Hotspot
Hotspot takes into consideration the spatial patterns and indicates the most affected areas. Identifying such areas would help on zeroing down into areas where most care has to be put and precise understanding of the type of crime can be gathered. Different types of hotspot detection techniques like Spatial Analysis Techniques, Interpolation Methods, Spatial Autocorrelation Methods.
- Spatial Analysis Methods
- Interpolation Methods
- Spatial Autocorrelation
Spatial analysis is the process of creating or extracting various information from spatial data, consisting of both locations and their attributes. Technique used under this category is Kernel Density Estimation.
Interpolation is the way of finding values of unknown points by making use of known points. These are not the original values, and are rather approximate ones. Interpolation is generally done in raster layers, but can be used in vector layers by making use of TIN. Interpolation methods used for the study is Inverse Distance Weighted Method.
Spatial Autocorrelation is the degree to which spatial features and their attributes are associated. It can be both positive and negative. Spatial Autocorrelation methods used for the study is Getis-Ord Gi*.
Study Area
The geographical coordinates of the study area lie between 3168327. 6 N to 3167690. 4 N and 718971 E to 720407. 1 E. The environments of all the data sets were set to the Geographic Coordinate System of Datum, WGS_1984, and Projected Coordinate System of Universal Transverse Mercator for Zone 43. In today day to day life the crime around us has significantly increased. There are different types of crimes such as burglary, theft, murder, etc. whose First Information Report (FIR) is filed. The officer has to look into the lists of records for analysis. This is time consuming and will result in the delay of action to be taken during Emergency situations. As the Aurangabad City is one of the fastest growing cities in Maharashtra, the crime rate is also increased in the past few years. As it is a Tourist spot, fraud cases and theft of personal belongings of the tourist also comes into the picture. The traditional methods used are not enough to cope up with the growing population and the crime.
The concept of Hot Spot has become an integral part of the study called crime analysis. A Hot Spot is a state of suggesting some arrangements of clusters in a spatial distribution. The clusters consisting of extremely high number of crime events within a geographical area are known as Hot Spots. Getting the knowledge of concentration of crime within an area, law enforcement agencies can enact timely and effective judgement of assigning police resources. In addition, Crime Hot Spots mapped by using GIS technology can admirably communicate crime patterns and crime prevention policies to decision makers and the public.
Results
- Remote Sensing Data
- Digitising Roads and Police Station
- Mapping different crime
- Analysing Results
Having setup the environment Hotspots were computed and found out for analysis.
5. Identifying hotspots
- Kernel Density Estimation Method
- Inverse Distance Weighted Method
- Getis-Ord Gi* Method
This method is used to find the density of cells in a raster by making use of known points and a kernel function, which is bi-variate in nature. It looks like a camel’s hump, having the known point at the centre as the peak, and the starting and ending points as 0. This method functions like a low pass filter and smoothens the areas. A KDE can be applied only on lines and polygons, and not on points.
This is a interpolation method that depends more on the nearer values for finding out the unknown value rather than the distant values. This creates shaded area, in form of layers. Unlike Kernal Density Estimation, this method works only on points.
This is a measure of global statistic, which is used to separate higher values from lower values. Unlike the Kernel Density Estimation Method and the Inverse Distance Weighted Method, this doesn’t have a smoothing or the layering effect.
The geographical coordinates of the study area lie between 3168327. 6 N to 3167690. 4 N and 718971 E to 720407. 1 E. The environments of all the data sets were set to the Geographic Coordinate System of Datum, WGS_1984, and Projected Coordinate System of Universal Transverse Mercator for Zone 43.
Roads have been digitised using lines, and police staions as points. They were properly labelled. Network analysis was to done to find out the nearest police station from the crime site, and to find out the best route from the police station to the scene.
Data of different crimes and their recorded cases were taken from the internet. For the purpose of experiment, hypothetical data was also used. The Crime of Murder, Day House Break and Night House Break has considered. Having done that, Crime Scenes in terms of points were plotted.
Conclusion
Various crimes are mapped and hotspots were found out using statistical tools. The study at hand will be of help to the police and other civic authorities to take necessary actions at the highly impacted areas and curb the problems. All the statistical tools used gives satisfactory results and are more precise than the other statistical tools. A KDE can be applied only on lines and polygons, and not on points and gives a smoothening effect. IDW creates shaded area, in form of layers and works only on points. GetisOrd Gi* works good with all vector datas and doesn’t have a smoothing or the layering effect. The result quality can be improved using real time and temporal data.
