Sensing chlorophyll, carotenoids and anthocyanin concentration in leaves with spatial resolution from digital image (invited manuscript)

CORDON, G. B.; LAGORIO, M. G.

In Color in Food: Technological and Psychophysical Aspects

CRC press

Lugar: New York; Año: 2012; p. 121 - 132

The aim of this work is to develop a computational vision-methodology to infer the pigments concentration in plant leaves with spatial resolution from a digital image. For this purpose, reflectance spectra of uniformly colored leaves were then recorded. On these leaves, chlorophyll-a, chlorophyll-b, carotenoids and anthocyanins were determined by extraction and posterior spectrophotometric measurement. From the reflectance spectra, the RGB coordinates were calculated. Then, correlations between the color coordinates and the pigment concentrations were obtained. On the other hand, images from heterogeneously colored leaves were captured using a commercial scanner. The scanned images were digitally processed to obtain R, G and B bands. Using the correlations between the colour coordinates (R, G and B) and pigments concentration, deducted for homogeneous leaves, the concentration of pigments per pixel in heterogeneous leaves were estimated