Design of a camera support and image analysis chain for rapid estimation of leaf area on de-tached leaves

Abstract

Leaf area is a central parameter in many plant science studies. It can be estimated non-destructively in situ using a proxy or remote sensing or destructively using a planimeter where each leaf has to be scanned. Non-destructive methods are generally faster but have greater biases and uncertainties. The planimeter approach, on the other hand, is more accurate but very time-consuming and requires specific, expensive equipment that is difficult to transport. In the context of monitoring leaf growth in potatoes, we have developed an intermediate approach to rapidly estimate the leaf area of detached leaves using quantitative imaging. We have designed a camera support with an image analysis chain coded in ImageJ and Python that can be used to measure leaf area under outdoor (field) or semi-controlled (shed, tunnel, greenhouse ....) conditions. Our method can be easily adapted to different plant species, at the level of canopy, plant or organ studies requiring rapid and inexpensive leaf area measurement. We present here the results of our system, with a sample of 683 plants collected and measured over 3 potato crop seasons. Validation was carried out with data acquired simultaneously with the planimeter, showing a very good correlation with our device and presenting a very interesting time saving, even if it slightly underestimates leaf area (17%).