Smart Image Storage and Retrieval For Crime Scene Investigation
Disha Rane, Bhavesh Shinde, Sidhesh Vaity, Priyanka Shingane · 2022 10th International Conference on Emerging Trends in Engineering and Technology - Signal and Information Processing (ICETET-SIP-22) · 2022
Crime scene investigation and initial evidence obtained from a crime scene plays a foundational role for solving any kind of criminal offense. As a result this paper proposes development of an efficient Crime scene investigation image storage and retrieval system which will emancipate law enforcement agencies from various technological and time constraints. The proposed system is developed with the help of various Transfer learning models incorporating algorithms of Residual Neural Networks such as ResNet18. Firstly the models are trained for categorization with the aim of storage by transfer learning model followed by models of ResNet18 for image retrieval. This paper proposes a system that can act as a repository of information regarding victims, suspects crime registration which can be used for comparative analysis, academic study and a tool for law enforcement policy making along with criminal investigation. Experimental Results obtained by testing the models on a custom-made crime scene investigation image data set have shown efficient performance by proposed algorithm