SonicVision: Navigating Object Counting Frontiers with Ultrasonic Economy
Mohsin Khalil, Mohsin Ikram, Adnan Shah · 2024
Precise enumeration of mobile entities, encompassing both individuals and objects, remains a pivotal necessity in numerous sectors including, but not limited to, transportation, surveillance, retail and industrial automation. In this work, we thoroughly discuss the limitations and practical uses of several technologies presently involved in the task of counting moving objects. Following a comparative analysis, we argue for the use of ultrasonic sensing technology and opine that it is the best-suited option for proximal object counting scenarios owing to its economic viability and superior durability. Consequently, we propose a microcontroller-centric framework Sonic Vision, thereby facilitating the streamlined development of an ultrasonic sensing based counting system. To substantiate our proposition, a prototype system has also been developed that effectively counts objects moving on a conveyor belt, thus underpinning the theoretical precepts with empirical evidence. Furthermore, with a prospective lens, we envisage potential realms of application, articulating scenarios where the said system might find utility, hence forging a pathway for future investigative endeavours in this domain.