A Human-Centered Approach For Tattoo Detection Using Convolutional Neural Networks: A Case Study In Forensic Applications

This paper explains the design, development and testing of a webbased system for detecting tattoos using Convolutional Neural Networks (CNNs) and a Human-Centered Design (HCD) approach. The system was created to help forensic professionals identify and analyze tattoos. The process included three main steps: collecting user feedback through surveys, creating wireframes for the design, and developing the system using deep learning techniques. User feedback was collected through surveys with forensic experts and general participants to understand their needs and expectations. Forensic professionals highlighted the importance of accuracy, secure data handling, and advanced filtering options, while general participants emphasized simplicity and ease of use. This input shaped the core features of the system, ensuring that it addressed both technical requirements and userfriendliness. Ethical concerns such as privacy, fairness, and transparency were also considered based on feedback received. The design process involved creating wireframes that served as visual blueprints for the system. These wireframes outlined the layout and functionality of key components, such as the dashboard and the tattoo analyzer. The dashboard was designed to provide quick access to search history, case management, and system metrics, while the tattoo analyzer allowed users to upload images or sketches for analysis. These designs were refined iteratively based on user feedback, ensuring that the final interface was intuitive and aligned with user needs.