Near-infrared imaging system for detecting small organic foreign substances in foods

Abstract

Contamination of foodstuffs with foreign substances is a serious problem because it often has negative effects on consumer health. However, detection of small organic substances in foods can be difficult because they are undetectable with traditional inspection apparatus. In this work, we developed new equipment that can detect small organic contaminant substances in food at high speed using a near-infrared (NIR) imaging technique. The absorption spectra of various foods were measured, and the spectra showed low absorbance at wavelengths from 600 nm to 1150 nm. Based on the observable wavelength range of a CMOS camera, which has a high dynamic range, superluminescent diodes (SLDs) with a wavelength of 830 nm were selected as light sources. We arranged 40 SLDs on a flat panel and placed a diffusion panel over them. As a result, uniformly distributed light with an intensity of 0.26 mW/cm² illuminated an area of 6.0 cm × 6.0 cm. Insects (3 mm wide) and hairs (0.1 mm in diameter) were embedded in stacked ham slices and in chocolate, with a total thickness of 5 mm in each case, and the transmission images were observed. Both insects and hairs were clearly observed as dark shadows with high contrast. We also compensated the images by using software developed in this study to eliminate low spatial frequency components in the images and improve the sharpness and contrast. As a result, the foreign substances were more clearly distinguished in the 5-mm-thick ham.