Year: 2025
Type: article
Abstract: Recently, image content based information retrieval has been widely used in many applications. A Mobile Visual Search (MVS) system involves capturing an image for the object to be searched for, extracting useful features, constructing object signature and searching local and/or remote databases for a match. This paper investigates the diverse effects of signature encryption on the accuracy, search burden and response time of an MVS system. The aim of this research is to provide a comprehensive understanding of the implications of signature encryption to figure out the design trade-offs that facilitate better MVS performance under given operation hypotheses. An MVS system employing different signature encryption algorithms is simulated with different system parameters. The obtained results show that signature encryption generally adds to system processing time from 5.9 µs to about 22.7 µs in test cases 4 and 6. Moreover, it makes the MVS system more sensitive to imaging and communication link imperfections leading to less capability to correctly identify entries and increased search and transmission loads from 1984 to 8075 requests to access the main database for test cases 3 and 9. These effects depend on the features of the specific encryption algorithm. Therefore, a trade-off between system performance and required security level should be considered to achieve a target performance suitable for the intended application
Source: Kufa Journal of Engineering
Institution: Northern Technical University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4415768124