The accurate geolocation of multimedia evidence is of critical importance to law enforcement and national security. While outdoor localization is well-supported by technologies such as GPS, reliable indoor positioning remains a significant challenge due to the frequent absence of discernible visual landmarks. This presentation addresses this gap by exploring the forensic use of Electrical Network Frequency (ENF), an artifact embedded in audiovisual recordings through interactions with the power grid. Existing research has predominantly utilized ENF for inter-grid classification, distinguishing recordings made in different continental power systems. This work, however, investigates the underexplored potential of subtle intra-grid ENF variations as a cue for location estimation within a single grid. A novel method for intra-grid localization is presented, based on ENF signals extracted from smartphone videos. The extraction of a clean ENF signal from smartphone recordings is a non-trivial task, complicated by significant sensor noise and aliasing artifacts.
To overcome this, we introduce an improved super-pixel-based technique for the robust isolation of the ENF signature. The proposed approach is validated using a dataset of real-world smartphone recordings collected from diverse locations within the Australian Eastern power grid. The extracted ENF signatures are compared against ground-truth data from grid anchor points, provided by a power supplier, demonstrating reliable intra-grid location estimation. To our knowledge, this work represents the first practical implementation achieving smartphone video localization in an intra-grid scenario.
Dr Li-minn Ang (Kenneth) received his BEng (Hons) and PhD degrees from Edith Cowan University in Australia. He is currently Professor of Electrical and Computer Engineering at the School of Science, Technology and Engineering at University of the Sunshine Coast. His research interests are in computer, electrical and systems engineering including intelligent systems and data analytics, machine learning, visual information processing, Internet of Things, embedded systems, wireless multimedia sensor systems, reconfigurable computing, and development of innovative technologies for real-world systems including smart cities, grids, engineering, agriculture, environment, health and defense. He is a Fellow of Engineers Australia.