Detection of russet on pome fruit using a spectral index (SRI)

Abstract

Russet is a brown necrotic appearance or physiological peel disorder of pome fruit. The objective of the present work was to identify russet in-situ on the fruit peel using new non-invasive technology. Colorimetry, spectrometry, 3D profilometry, and a luster sensor were employed on green ‘Conference’ and red ‘Williams Christ’ pears. Non-invasive true and false color images at 40x magnification with a 3D-profilometer type VR-3000 (Keyence) allowed identification of russet on ‘Conference’ pear with a larger roughness Ra 3.3 µm (SE 0.4) compared with Ra 2.5 µm (SE 0.3) on russet-devoid fruit peel. The contactless russet detection using the luster sensor CZ-H72 failed to detect a significant difference (8 a.u.) in glossiness between russet (54.5 a.u.) and nonrusset (62.5 a.u.) on ‘Conference’ (green) in contrast to red ‘Williams Christ’ with a ca.2fold statistically significant difference between russet (38.4 a.u) and non-russet (67.5. a.u.) surface. The situation reversed with colorimetry: For the red pears of ’Williams Christ’, differentiation between russet (66.8 °hue) and russet-devoid peel (62.4 °hue) was impossible when using contactless colorimetry. Russet could be successfully detected for all cultivars examined using portable contactless spectrometry (190-1100 nm). A new spectral russet index (SRI) was proposed relating the two peaks of light reflection (550-600 nm and 775-785 nm) to the trough (667-685 nm). Russeted peel (e.g., 3.6) exhibited ca. 25% smaller SRI values than russet-devoid peel (4.8) of both cultivars ‘Conference’ (green) and ‘Sweet Sensation’(red). Overall, both portable noninvasive techniques, 3D-profilometer and spectral light reflection with the novel russet index, proved suitable for real-time, in-situ russet detection in pears. In particular, the study has identified colorimetry as a suitable measure for russet detection for green and luster sensors for red pear cultivars, which offers new possibilities for contactless russet detection in the field or on a grading line.

Publication