Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297

A Novel Method for Distance Calculation from

Forensic Sketches Converted from Images
Shreyesh Sanjeev1; Litty Jose2
2
HOD; Assistant Professor
1,2
Department of Forensic Science, St Thomas College, Thrissur

Publication Date: 2025/05/02

Abstract: Accurate crime scene documentation is essential in forensic investigations, yet traditional sketching methods are
time-consuming, prone to human error, and reliant on investigator skill. This study presents a forensic application designed
to automate the sketching and measurement process, enhancing efficiency and precision. The application utilizes Sobel edge
detection to convert crime scene photographs into sketches, allowing threshold adjustments for optimal detail
representation. Additionally, it incorporates a measurement system that computes distances between evidence points based
on four fixed reference points. By inputting known straight-line and diagonal distances, investigators can obtain precise
spatial measurements without manual calculations. Tested across fifteen simulated crime scenes, the software successfully
validated two hypotheses: (1) edge detection can generate detailed forensic sketches, and (2) accurate measurements can be
computed using fixed reference points. The model demonstrated 90-95 percentage accuracy, though challenges such as image
warping and background noise require further refinement. This study underscores the potential of digital forensic tools in
modern investigations. While the prototype is functional, future enhancements including AI integration, advanced noise
reduction, and improved security could further optimize its reliability. Once fully developed, this application has the
potential to standardize digital crime scene documentation, improve forensic reconstructions, and provide a valuable
resource for law enforcement and legal professionals.

Key Words: Image to Sketch, Sobel Operator, Edge Detection, Flutter, Crime Scene Documentation.

How to Cite: Shreyesh Sanjeev; Litty Jose (2025), A Novel Method for Distance Calculation from Forensic Sketches Converted
from Images. International Journal of Innovative Science and Research Technology, 10(4), 2093-2098.
https://doi.org/10.38124/ijisrt/25apr1297

I. INTRODUCTION  Objectives

Accurate crime scene documentation is crucial in  Develop a working prototype of an app using flutter
forensic investigations, yet traditional sketching methods are software that converts images into sketches using edge
time-consuming, error-prone, and require specialized skills. detection
This project introduces a forensic application that automates  Implement an option to adjust the threshold using a slider
crime scene sketching by converting photographs into to control the level of detail included in the sketch.
detailed sketches using Sobel edge detection while  Integrate a method to compute the measurements of
incorporating a measurement tool that calculates distances evidence in the sketch relative to fixed reference points by
based on four fixed reference points. By integrating digital allowing users to input the distance between them.
forensic techniques, the software enhances efficiency and
accuracy in crime scene reconstruction, reducing reliance on  Hypothesis
manual methods. Tested on multiple simulated crime scenes,
it demonstrated an accuracy rate of approximately 90-95%,  A detailed fair sketch can be generated from an image
though challenges such as image distortion and background using edge detection, with adjustable threshold settings to
noise remain. As forensic science evolves, adopting refine the level of detail.
automated tools can standardize crime scene documentation,  Accurate measurements of an object’s distance in the
improve reliability, and streamline investigations. Future sketch can be computed if four fixed reference points are
enhancements, including AI integration and advanced noise provided and the distances between them are inputted.
reduction, will further refine the software, making it a
valuable resource for forensic professionals and law
enforcement agencies.

IJISRT25APR1297 www.ijisrt.com 2093

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297
II. METHODOLOGY allowing the software to compute measurements for other
evidence points automatically. The accuracy of these
This study aimed to develop a forensic application that computed measurements was assessed by comparing them
automates crime scene sketching and measurement to with manually recorded values.
improve accuracy and efficiency in forensic documentation.
The results demonstrated a 90-95% accuracy rate,
The procedure involved capturing photographs of confirming the feasibility of the approach while highlighting
fifteen simulated crime scenes, ensuring proper lighting and challenges such as image distortion and background noise.
a clear bird’s-eye or 90-degree perspective. These images Future refinements, including AI integration, advanced noise
were uploaded into the application, where they were reduction, and improved measurement calibration, are
processed into sketches. Users then selected four fixed proposed to enhance the software’s reliability and forensic
reference points and inputted known distances between them, applicability.

III. RESULTS

Fig 1 Indoor Crime Scene (Homicide)

IJISRT25APR1297 www.ijisrt.com 2094

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297

Fig 2 The Fixed Points are ‘A’&’B’ Corners of the Room. ‘C’ Edge of Bed. ‘D’ Middle of Sofa

Fig 3 Indoor Crime Scene (Knife Attack)

IJISRT25APR1297 www.ijisrt.com 2095

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297

Fig 4 Fixed Points ‘A’ Middle of Table. ‘B’ Corner of Room ‘C’ Corner of Frame ‘D’ Corner of Room

Fig 5 Outdoor Crime Scene (Suicide)

IJISRT25APR1297 www.ijisrt.com 2096

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297

Fig 6 Fixed Points ‘A’&’B’ Edge of Wall. ‘C’ Tree Stump ‘D’ Rock

IV. HYPOTHEIS-BASED INTEPRETATION documentation by reducing manual errors and improving

efficiency. Future improvements, such as advanced
The study successfully validated both hypotheses. The calibration techniques and noise reduction methods, could
first hypothesis stated that a detailed forensic sketch could be further optimize the accuracy and reliability of the software.
generated from an image using edge detection with adjustable
threshold settings. This was confirmed through the V. DISCUISSION AND CONCLUSION
application’s use of Sobel edge detection, which effectively
converted crime scene photographs into sketches while  Major Findings
allowing users to refine details. This study successfully developed a forensic
application that automates crime scene sketching and
The second hypothesis proposed that accurate measurement, reducing reliance on manual methods. The
measurements of an object’s distance within the sketch could software converts photographs into sketches using Sobel edge
be computed if four fixed reference points were provided. detection and calculates distances between evidence points
Testing across multiple simulated crime scenes confirmed based on four fixed reference points. Tested across fifteen
this, with the software achieving 90-95% accuracy in distance simulated crime scenes, the application demonstrated an
calculations. However, minor discrepancies due to image accuracy rate of approximately 90-95%, validating its
distortion and background noise highlighted the need for effectiveness. The results confirmed that edge detection can
further refinement in measurement precision. generate detailed forensic sketches and that precise
measurements can be computed when fixed reference points
Overall, the findings demonstrate that automated are provided, making crime scene documentation more
sketching and measurement tools can enhance forensic efficient and reliable.

IJISRT25APR1297 www.ijisrt.com 2097

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr1297
 Limitations [8] LoGrande, Marc (2016) "The Utilization of Mobile
While the software performed well, certain challenges Technology for Crime Scene Investigation in the San
were identified. Image distortion and background noise Francisco Bay Area," Themis: Research Journal of
impacted measurement accuracy, and the system requires Justice
properly captured images taken from a bird’s-eye or 90- StudiesandForensicScience:Vol.4,Article9.https://doi.
degree perspective for optimal results. Additionally, the lack org/10.31979/THEMIS.2016.0409
of built-in image-saving functionality means users must take https://scholarworks.sjsu.edu/themis/vol4/iss1/9
screenshots, which may affect image quality. [9] Mathur, Neha & Mathur, Shruti & Mathur, Divya.
(2016). A Novel Approach to Improve Sobel Edge
 Suggestions for Further Studies Detector. Procedia Computer Science. 93. 431-438.
Future research should explore integrating multiple 10.1016/j.procs.2016.07.230.
edge detection methods, such as Canny, to improve [10] Moulay Abdelmajid Kassem., Khalid Lodhi (2024)
adaptability based on image quality. Noise reduction Revolutionizing Forensic Science: The Role of
techniques like Gaussian blur could enhance clarity, while Artificial Intelligence and Machine Learning
security features such as encryption and tamper-proofing Intelligence and Machine Learning International
could ensure forensic reliability. Expanding the application Journal of Data Science 2024, pp. 10–
for mobile platforms would enhance accessibility, allowing 18https://doi.org/10.5147/ijds.vi.255
investigators to document crime scenes more effectively in [11] Mutneja, Vikram. (2015). Methods of Image Edge
real time. Detection: A Review. Journal of Electrical &
Electronic Systems. DOI 04. 10.4172/2332-
 Implications 0796.1000150.
Automating crime scene sketching and measurement [12] Peter W. PFEFFERLI (2005) COMPUTER AIDED
has significant implications for forensic science. CRIME SCENE SKETCHING Problems of Forensic
Standardizing documentation minimizes human error, Sciences, vol. XLVI, 2001, 83–85 Received 4 May
improves investigative efficiency, and strengthens legal 2005; accepted 15 September 2005
proceedings. With further refinements, this technology could [13] Raghav Brahmadesam Venkataramaiyer (2022)
become a valuable tool for forensic professionals, Geometric Understanding of Sketches eprint
transforming the way crime scenes are documented and "2204.06675 https://arxiv.org/abs/2204.06675
analysed. [14] Saikia, Surajit & Fidalgo, Eduardo & Alegre, Enrique
& Fernández-Robles, Laura. (2017). Object Detection
REFERENCES for Crime Scene Evidence Analysis Using Deep
Learning. 14-24. 10.1007/978-3-319-68548-9_2.
[1] Aneglika Dudek, Anna Dąbek, Iwona Zborowska, [15] Sharma, Bhoopesh & Bashir, Raeesa & Philip, Sharon
Jakub Lichosik (2023) Integrating artificial & Kumar, Hardeep. (2019). A Comparative Study of
intelligence in forensic science Issue Year: 10/2023 Mobile Applications for Crime Scene Measurements-
Issue No: 10 Page Range: 13-31 Page Count: 18 A Digital Approach. 492-495.
[2] Arifi, Besim. (2015). Documentation of the Crime 10.1109/ICCIKE47802.2019.9004348.
Scene. European Journal of Interdisciplinary Studies. [16] Tondare, Balram & Shirsat, Rohan & Tiwari, Aditya &
1. 32. 10.26417/ejis.v1i2.p32-38. Kute, Kalyani & Bhagat, Abhilasha. (2024). Forensic
[3] Bhukya, Teja, An Overview of Crime Scene Sketching Sketch to Real Image. 2024 IJNRD | Volume 9, Issue
(2023)http://dx.doi.org/10.2139/ssrn.4461745 5 May 2024| ISSN: 2456-4184 | IJNRD.ORG
[4] Chango, X., Flor-Unda, O., Gil-Jiménez, P., & Gómez-
Moreno, H. (2024). Technology in Forensic Sciences:
Innovation and Precision. Technologies, 12(8), 120.
https://doi.org/10.3390/technologies12080120
[5] Davis, M., Compton, S.S., Borovicka, N., Castle, S.D.,
Rushton, C.G., & Staton, P.J. (2014). Mobile Crime
Scene Applications : An Evaluation of Their Use and
Future Direction MUFSC, 1401 Forensic Science Dr.
Huntington, WV 25701
[6] Han, Lili & Tian, Yimin & Qi, Qianhui. (2020).
Research on edge detection algorithm based on
improved sobel operator. MATEC Web of
Conferences. 309. 03031.
10.1051/matecconf/202030903031.
[7] Junfeng Jing, Shenjuan Liu, Gang Wang, Weichuan
Zhang, Changming Sun, Recent advances on image
edge detection: A comprehensive review,
Neurocomputing, Volume 503, 2022, Pages 259-271,
ISSN 0925-2312,
https://doi.org/10.1016/j.neucom.2022.06.083.

IJISRT25APR1297 www.ijisrt.com 2098