Thumbnail Image

Thesis

Browse

2017 - 20194
 Reset filters

Settings

Subject: Forensic Science ×

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    ThesisItemOpen Access
    Multiple Classifier System in Offline Handwriting for Writer Identification and Verification
    (Department of Forensic Science Sam Higginbottom University of Agriculture, Technology and Sciences Allahabad, U.P., India, 2018) Nigam, Ms. Kriti; Mishra, Dr. Munish Kumar
    The study entitled “Multiple Classifier System in Offline Handwriting for Writer Identification and Verification” was carried out at the Department of Forensic Science, SHUATS, Allahabad with the following major objectives-1) To study the existing handwriting classifier systems for writer identification and verification; 2) To develop a classifier addressing the drawbacks in existing systems; and 3) To analyze the system developed using standard techniques and methodologies For conducting this study, handwriting samples of 400 writers were randomly collected from different regions of Uttar Pradesh of age range between 18 to 60 years, without letting them know the purpose of study. The participants were requested to write the control passage dictated to them in the unlined space of the same make of paper using the pen provided. In the first part of this research work, a multiple classifier system was developed i.e those characteristics were evaluated in the writing samples which can significantly differentiate handedness, gender and medium of education of writer. Different handwriting characteristics were then studied which were specific to handedness, gender and medium of education of the writer. Observed characteristics were then statistically evaluated using the Chi-Square method. In second phase, writer verification was done by matching the class and individual characteristics of two different writing samples to determine whether they belong to the same source or not. In this part various handwriting features like area, perimeter, width, height, angle of various letters in the selected words i.e. have, you, will, give and money, were calculated mathematically using Digimizer- Image Analysis Software. Based on the results of above work, an automated tool for handwriting examination was developed using Visual Studio 2010, to ease out the process of handwriting analysis and to bring a significant change in the field of questioned document examination. The developed automated tool works in a sequential manner, where the user can upload the questioned and specimen images. The first step of the tool is comprised of handedness based classifier. It offers the user a set of questions regarding the presence or absence of the observed characteristics which were statistically found to be significant in differentiating between the probable handedness i.e sinistrality or dextrality of the writer. Each characteristic has a fixed probability value associated with it. The overall result will give the net probability percentage of writer being sinistral or dextral, which will also be represented by a pie diagram. The second and third steps of the tool are comprised of gender and medium of education based classifiers respectively. Fourth step of the tool deals with the comparison of different class characteristics of the questioned and specimen writings. In this step, the tool offers six class characteristics, each with its different variants. User has to choose the correct variant of each characteristic for both questioned and specimen writing. The tool will give the result in form of match or mismatch of the characteristics, which will also be represented by a line diagram. The last step of the tool deals with the comparison of the metric values i.e. angle of the highest point of the first letter, angle of the highest point of the second letter, area, perimeter, width and height of the selected words of the questioned and specimen writings. In this step, the user has to input these metric values, the tool will compare the input value with the database of measurements and will display the name of the writer with which the measurements are matching most. Finally the system so developed was compared with the existing systems, addressing their limitations. In the proposed method, a mathematical approach has been made based on metric variations between handwriting samples obtained from different writers to provide assistance to the standing opinion of Examiners of Document so as to provide a mathematical support for every identified feature, which in manual examination is merely based on perception.
  • Thumbnail Image
    ThesisItemOpen Access
    Re-creation of crime scene usingDifferential Global Positioning System (DGPS) in outdoor simulated scene.
    (Department of Forensic Science Sam Higginbottom University of Agriculture, Technology and Sciences Allahabad-211007 (U.P.) India, 2018) KENNAO, PULENO; Kesharwani, Dr. Lav
    It is vital to envisage the location of the evidence to get an idea on how the crime took place and in what manner. Managing outdoor crime scenes with the application of conventional method can be a demanding job to achieve exclusively when the crime emerged in big space, where there is lack of reference points and tress being surrounded. In this research, fifteen simulated outdoor crime scenes were created, where both manual tape measurement and DGPS unit was applied to see if the differential GPS offers an accurate and reliable alternative for mapping the scene over the baseline method. To develop a final digital map of the simulated outdoor scenes, Geographical Information System (GIS) was utilized. The result indicates that the DGPS unit provides a consistent reading when compared to manual tape measurement giving an average reading difference of 0.1, 0.2, 0.12, 0.13, 0.01, 0.2, 0.06, 0.04, 0.2, 0.2, 0.3, 0.1, 0.1, 0.06, (inch) from fourteen simulated outdoor crime scenes. The study concluded that DGPS unit can be a viable alternative for mapping the outdoor crime scenes. With the population exceeding 1 million people, the Allahabad city is in the ten most populous cities of the state and highest crime recorded according NCRB. It is one of the fastest developing cities of the area. This research also studies the spatial distribution of Police Station in each zone of Allahabad city based on crime and population in corporation with Differential Global Positioning System (DGPS) by visiting each Police Station to get the exact locations (latitude and longitude) while the primary data, secondary data and the demographic data were all sourced from SSP, Police Station and Municipal Cooperation, Allahabad. The data were analyzed using simple descriptive, pie chart and frequency distribution table. ArcGIS 10.4 software was used to draw the map of each zone showing the presence of Police Station. Chi-square test has shown that the Police Station in each zone is not evenly distributed. It was also discovered through 10 | P a g e Coefficient Correlation and T-test that as population increase, crime also increases. This research concludes that crime increases with tremendous growth in population, and there is a need for the population consideration in deployment of police stations in the concerned zones for achieving better security situations in the city. This study highlights the importance of identifying the most reported crime in the police stations of Allahabad city through the thematic map that can be utilized by the police departments. GIS can be used as a tool to identify factors contributing to crime, and thus allow police to proactively respond to the situation before they show a red signal. Hotspot map was determined through interpolation method (IDW) to see the concentration of different crime in each police station. The crime challenges faced by the police department need to pursue to implement computerizes crime mapping for crime analysis. The result shows that the jurisdiction of Doomanganj Police Station and Kernal Ganj Police Station needs high patrolling and effective measures for the type of crime being conducted in this study
  • Thumbnail Image
    ThesisItemOpen Access
    ANALYSIS OF RARELY DISGUISED FEATURES IN HANDWRITINGS FOR FORENSIC CONSIDERATION
    (DEPARTMENT OF FORENSIC SCIENCE SAM HIGGINBOTTOM UNIVERSITY OF AGRICULTURE, TECHNOLOGY AND SCIENCES PRAYAGRAJ, U.P., INDIA, 2019) HARNE, PRAJAKTA; Mishra, Dr. Munish Kumar
    The proposed study is based on the application of statistical approach in the science of handwriting identification for the analysis of disguised pattern. Disguised handwritings are those handwritings where the author attempts to intentionally conceal his natural writing habits. Disguised writing does not have the spontaneous appearance of naturally executed words and letters. The exercise of obscuring one’s natural writing habits while, at the same time, writing in an entirely different fashion entails extraordinary attentiveness and determination if the deception is to avoid detection. The most common method used by the individuals while providing specimens to the court is to disguise their handwriting samples and similar means of disguise are being used in anonymous letter where individuals try to hide their identity. Thus, the proposed study is intended to emphasize those features of handwriting which are rarely disguised. These features not only constitute class and individual characteristics of handwriting, but also incorporate natural variations so as to impart further individualization to the calligraphic script. For implementing the project, a large number of samples are taken from volunteers of different age groups, diverse professional backgrounds, gender and geographical regions. These also include matured writers, as well as individuals who are routinely not doing writing work. The application of statistical methods estimates the percentage of various parameters chosen for study. During analysis of sample, 13 parameters of handwriting were observed to be rarely disguised. It has also been observed that writers usually alter their overall design of letters without altering the identified disguise parameters. The objective of the present study is also placing stress on the fact that, if this set of parameters remains unchanged while executing disguise writings then it infers that they can be considered as factors of identification the common authorship which is the fundamental motive behind every examination carried out in handwriting identification. The proposed research work is supplemented with graphs indicating the number of individuals using the given parameter in different manner.
  • Thumbnail Image
    ThesisItemOpen Access
    ESTABLISHMENT OF GENUINESS OF HANDWRITING/SIGNATURE BASED ON SEQUENCE OF STROKES & OTHER INDIVIDUAL CHARACTERS
    (Department of Forensic Science Sam HigginbottomInstitute of Agriculture, Technology & Sciences (Formerly Allahabad Agricultural Institute- Deemed-to-be-University) Allahabad-211007, India, 2017) KUMAR, SUNEET; Vaid, B. A.
    The present study entitled “Establishment of Genuineness of Handwriting/Signature based on sequence of strokes & other individual characters” was conducted in the Department of Forensic Science (SHIATS, Allahabad) for the preparation of samples to analyze the samples of Writings/Signatures. The verification of handwritten signatures is one of the oldest and the most popular biometric authentication methods in our society. A history which spans several hundred years has ensured that it also has a wide legal acceptance all around the world. This thesis addresses the various problems related to handwriting/signatures identification. Most of the writer identification work is being done manually till date as a lot of context dependent information, such as, source of documents, nature of handwriting, etc. is difficult to model mathematically. However, they can be easily analyzed by human experts. Still, an automatic handwriting analysis system is useful as it can remove subjectivity from the process of hand- writing identification and can be used for expert advice in various court cases. The present study carried out at the school of forensic science, SHIATS, Allahabad, deals with signature samples, collected for Verification on the basis of strokes and other characteristics.. In the present study an attempt is made to provide an effective method of identification of the imitated signature on the basis of their sequence of strokes are as one of the most important feature, for the verification of a genuine signature, and none of the other dynamic information can be recovered without the strokes. This work is focused on the recovery of stroke and other characteristics like slant, alignment, vertical height, horizontal length, no of pen-down and cursiveness. Handwriting is neuromuscular co-ordination which is made up of various elements, some of these elements can be seen easily and some of them are inconspicuous but very significant. Handwriting has the quality of roundness, smoothness, legibility or speed. Handwriting is identified by a study and comparison of its quality, elements, features, characteristics. All these various consequences of handwriting vary greatly as evidence of identity or difference. These characteristics may be changed due to many reasons, which are writing materials, temporary physical or physiological deviations, natural variation, age etc. In the present study handwriting samples of different persons are examined, for the variation in handwriting, their types and reasons for their deviation beyond the range of variation is described on the basis of experimental results observed.