Smartphone Colorimetry with In-Frame Color Calibration: Hue-Based Discrimination of Diesel-Type Fuels and Ethanol–Gasoline Blends
Year: 2026
Type: preprint
Abstract: Color measurements from consumer cameras are difficult to compare because device-specific image processing is variable and often undocumented. We present a fully smartphone-based workflow, in which the sample and a color reference are captured simultaneously, enabling image-specific correction. Nile Red-based test strips were measured in a compact PhotoBox with an integrated ColorChecker chart. A color-correction matrix derived individually for each image reduced the mean hue error across the chromatic range of the strips from 5.1 to 1.6°. Using calibrated hue angle as the sole analytical parameter, eight fuels and ethanol–gasoline blends formed largely distinct clusters, with a mean within-fuel scatter of 1.5°. The regulatory-relevant nominal E10 and E20 reference samples were separated by 14.2°. Within the reference dataset, leave-one-strip-out cross-validation correctly assigned 46 of 48 strips (95.8%); both misassignments were from E5 to E10 clusters. Between-batch differences were ≤0.6°, and the inter-device standard deviation for a fixed paper target was 2.0°. Two nominally identical E10 fuels differing in matrix composition deviated by approximately 12°, suggesting that hue reflects effective fuel polarity rather than ethanol content alone. This compact workflow enables reproducible fuel discrimination using a consumer smartphone. A blind validation using unknown samples has yet to be conducted.
Source: Preprints.org
Language: English
Subfield: Analytical Chemistry
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225114247