A Novel Multipurpose Snapshot Hyperspectral Imager used to Verify Security Hologram
Arvind Mukundan, Hsiang‐Chen Wang, Yu-Ming Tsao
Abstract
In this study, a Low-cost and portable module is built using a Raspberry Pi camera and microprocessor that captures the images of security holograms which are fed into the hyperspectral imaging algorithm that measures the reflectance to differentiate the counterfeit from the original hologram. Reflectivity is one of the difficulties in distinguishing a duplicate hologram from an authentic one. Due to the fact that a little change in lighting situation would entirely alter the hologram’s reflection pattern, a standardized detector for duplicate holograms has not yet been developed. In this work, hyperspectral imaging (HSI)-based housing module for distinguishing between authentic and counterfeit holograms was presented. The module included a Raspberry Pi 4 CPU, a Raspberry Pi camera, an LCD display, and an LED lighting system with a dimmer. A visible HSI algorithm capable of transforming a Raspberry Pi camera’s RGB picture into a hyperspectral image was developed. Mean gray value (MGV) and reflectivity measurements were performed on an area of interest (ROI) extracted from the spectral picture. Results indicated that shorter wavelengths are optimal for distinguishing holograms when using MGV as the classification criterion, but longer wavelengths are optimal when utilizing reflectivity. This design is distinguished by its cheap cost, simplicity, absence of moving components, and lack of need for an extra decoding key.